3.1 属性权重为直觉模糊数的加权直觉模糊数
属性
G
j
∈
G
G_{j}\in G
Gj∈G的权重为直觉模糊数
ω
~
j
=
⟨
ρ
j
,
τ
j
⟩
(
j
=
1
,
2
,
.
.
.
,
n
)
\tilde{\omega}_{j} = \left\langle\rho_{j},\tau_{j}\right\rangle(j=1,2,...,n)
ω~j=⟨ρj,τj⟩(j=1,2,...,n)满足条件
ρ
j
∈
[
0
,
1
]
\rho_{j}\in[0,1]
ρj∈[0,1]、
τ
j
∈
[
0
,
1
]
\tau_{j}\in[0,1]
τj∈[0,1],且
0
≤
ρ
j
+
τ
j
≤
1
0\leq\rho_{j}+\tau_{j}\leq1
0≤ρj+τj≤1,则
F
~
i
j
\tilde{F}_{ij}
F~ij的加权直觉模糊数为
F
′
~
i
j
=
⟨
μ
i
j
′
,
ν
i
j
′
⟩
‾
=
ω
~
j
⊗
F
′
~
i
j
=
⟨
ρ
j
,
τ
j
⟩
⟨
μ
i
j
,
ν
i
j
⟩
=
⟨
ρ
j
μ
i
j
,
τ
j
+
ν
i
j
−
τ
j
ν
i
j
⟩
‾
(3.1)
\begin{aligned} \tilde{F^{'}}_{ij} &= \color{red}{\underline{\left\langle\mu^{'}_{ij},\nu^{'}_{ij}\right\rangle}} \\ &= \tilde{\omega}_{j}\otimes\tilde{F^{'}}_{ij} \\ &= \left\langle\rho_{j},\tau_{j}\right\rangle\left\langle\mu_{ij},\nu_{ij}\right\rangle \\ &= \color{red}{\underline{\left\langle\rho_{j}\mu_{ij},\tau_{j}+\nu_{ij}-\tau_{j}\nu_{ij}\right\rangle}} \tag{3.1} \end{aligned}
F′~ij=⟨μij′,νij′⟩=ω~j⊗F′~ij=⟨ρj,τj⟩⟨μij,νij⟩=⟨ρjμij,τj+νij−τjνij⟩(3.1)
方案
Y
i
(
1
,
2
,
.
.
.
,
m
)
Y_{i}(1,2,...,m)
Yi(1,2,...,m)的加权直觉模糊综合评价结果为
d
~
i
=
ω
~
1
⊗
F
′
~
i
1
⊕
ω
~
2
⊗
F
′
~
i
2
⊕
.
.
.
⊕
ω
~
n
⊗
F
′
~
i
n
=
⟨
ρ
1
μ
i
1
,
τ
1
+
ν
i
1
−
τ
1
ν
i
1
⟩
+
⟨
ρ
2
μ
i
2
,
τ
2
+
ν
i
2
−
τ
2
ν
i
2
⟩
+
.
.
.
+
⟨
ρ
n
μ
i
n
,
τ
n
+
ν
i
n
−
τ
n
ν
i
n
⟩
=
∑
j
=
1
n
⟨
μ
i
j
′
,
ν
i
j
′
⟩
\begin{aligned} \tilde{d}_{i} &= \tilde{\omega}_{1} \otimes \tilde{F^{'}}_{i1} \oplus \tilde{\omega}_{2} \otimes \tilde{F^{'}}_{i2} \oplus ... \oplus \tilde{\omega}_{n} \otimes \tilde{F^{'}}_{in}\\ &= \left\langle\rho_{1}\mu_{i1},\tau_{1}+\nu_{i1}-\tau_{1}\nu_{i1}\right\rangle + \left\langle\rho_{2}\mu_{i2},\tau_{2}+\nu_{i2}-\tau_{2}\nu_{i2}\right\rangle + ... + \left\langle\rho_{n}\mu_{in},\tau_{n}+\nu_{in}-\tau_{n}\nu_{in}\right\rangle \\ &= \sum_{j=1}^{n} \left\langle\mu^{'}_{ij},\nu^{'}_{ij}\right\rangle \end{aligned}
d~i=ω~1⊗F′~i1⊕ω~2⊗F′~i2⊕...⊕ω~n⊗F′~in=⟨ρ1μi1,τ1+νi1−τ1νi1⟩+⟨ρ2μi2,τ2+νi2−τ2νi2⟩+...+⟨ρnμin,τn+νin−τnνin⟩=j=1∑n⟨μij′,νij′⟩
那么,
d
~
i
=
⟨
μ
i
′
,
ν
i
′
⟩
=
∑
j
=
1
n
⟨
μ
i
j
′
,
ν
i
j
′
⟩
=
⟨
1
−
(
1
−
μ
i
1
′
)
(
1
−
μ
i
2
′
)
.
.
.
(
1
−
μ
i
n
′
)
,
ν
i
1
′
ν
i
2
′
.
.
.
ν
i
n
′
⟩
‾
(3.2)
\color{red} { \begin{aligned} \tilde{d}_{i} &= \left\langle\mu^{'}_{i},\nu^{'}_{i}\right\rangle \\ &= \sum_{j=1}^{n} \left\langle\mu^{'}_{ij},\nu^{'}_{ij}\right\rangle \\ &= \color{red}{\underline{\left\langle1-(1-\mu^{'}_{i1})(1-\mu^{'}_{i2})...(1-\mu^{'}_{in}),\nu^{'}_{i1}\nu^{'}_{i2}...\nu^{'}_{in}\right\rangle}} \tag{3.2} \end{aligned} }
d~i=⟨μi′,νi′⟩=j=1∑n⟨μij′,νij′⟩=⟨1−(1−μi1′)(1−μi2′)...(1−μin′),νi1′νi2′...νin′⟩(3.2)
方案
Y
i
(
i
=
1
,
2
,
.
.
.
,
m
)
Y_{i}\left(i=1,2,...,m\right)
Yi(i=1,2,...,m)的加权综合评价结果
d
~
i
\tilde{d}_{i}
d~i的得分值和精确值为:
s
(
d
~
i
)
=
μ
i
′
−
ν
i
′
(
i
=
1
,
2
,
.
.
.
,
m
)
,
h
(
d
~
i
)
=
μ
i
′
+
ν
i
′
(
i
=
1
,
2
,
.
.
.
,
m
)
s\left(\tilde{d}_{i}\right) = \mu^{'}_{i}-\nu^{'}_{i}\left(i=1,2,...,m\right),h\left(\tilde{d}_{i}\right) = \mu^{'}_{i}+\nu^{'}_{i}\left(i=1,2,...,m\right)
s(d~i)=μi′−νi′(i=1,2,...,m),h(d~i)=μi′+νi′(i=1,2,...,m)
利用直觉模糊数排序规则,可以确定方案
Y
i
(
i
=
1
,
2
,
.
.
.
,
m
)
Y_{i}\left(i=1,2,...,m\right)
Yi(i=1,2,...,m)的优劣次序。
3.2 属性权重为直觉模糊数的多属性决策方法分析步骤
属性权重为直觉模糊数的多属性决策方法的步骤如下:
步骤S1 确定多属性决策问题的方案集
Y
=
{
Y
1
,
Y
2
,
.
.
.
,
Y
m
}
Y=\left\{Y_{1},Y_{2},...,Y_{m}\right\}
Y={Y1,Y2,...,Ym}和属性集
G
=
{
G
1
,
G
2
,
.
.
.
,
G
n
}
G=\left\{G_{1},G_{2},...,G_{n}\right\}
G={G1,G2,...,Gn};
步骤S2 获取多属性决策问题中方案
Y
i
∈
Y
Y_{i} \in Y
Yi∈Y关于属性
G
j
∈
G
G_{j} \in G
Gj∈G的直觉模糊特征信息,构建直觉模糊决策矩阵
F
F
F;
步骤S3 确定多属性决策问题各属性的直觉模糊权重,得到属性的直觉模糊权重向量
ω
~
=
(
ω
~
1
,
ω
~
2
,
.
.
.
,
ω
~
n
)
T
=
(
⟨
ρ
1
,
τ
1
⟩
,
⟨
ρ
2
,
τ
2
⟩
,
.
.
.
,
⟨
ρ
n
,
τ
n
⟩
)
T
\tilde{\omega} = {\left(\tilde{\omega}_{1},\tilde{\omega}_{2},...,\tilde{\omega}_{n}\right)}^{T}={\left(\left\langle\rho_{1},\tau_{1}\right\rangle,\left\langle\rho_{2},\tau_{2}\right\rangle,...,\left\langle\rho_{n},\tau_{n}\right\rangle\right)}^{T}
ω~=(ω~1,ω~2,...,ω~n)T=(⟨ρ1,τ1⟩,⟨ρ2,τ2⟩,...,⟨ρn,τn⟩)T;
步骤S4 利用式
F
′
~
i
j
=
⟨
μ
i
j
′
,
ν
i
j
′
⟩
=
ω
~
j
⊗
F
′
~
i
j
=
⟨
ρ
j
,
τ
j
⟩
⟨
μ
i
j
,
ν
i
j
⟩
=
⟨
ρ
j
μ
i
j
,
τ
j
+
ν
i
j
−
τ
j
ν
i
j
⟩
\tilde{F^{'}}_{ij} = \left\langle\mu^{'}_{ij},\nu^{'}_{ij}\right\rangle=\tilde{\omega}_{j}\otimes\tilde{F^{'}}_{ij}= \left\langle\rho_{j},\tau_{j}\right\rangle\left\langle\mu_{ij},\nu_{ij}\right\rangle=\left\langle\rho_{j}\mu_{ij},\tau_{j}+\nu_{ij}-\tau_{j}\nu_{ij}\right\rangle
F′~ij=⟨μij′,νij′⟩=ω~j⊗F′~ij=⟨ρj,τj⟩⟨μij,νij⟩=⟨ρjμij,τj+νij−τjνij⟩计算多属性决策问题的加权直觉模糊决策矩阵
F
′
=
(
F
′
~
i
j
)
m
×
n
F^{'}={\left(\tilde{F^{'}}_{ij}\right)}_{m×n}
F′=(F′~ij)m×n;
步骤S5 利用式
d
~
i
=
⟨
μ
i
′
,
ν
i
′
⟩
=
∑
j
=
1
n
⟨
μ
i
j
′
,
ν
i
j
′
⟩
=
⟨
1
−
(
1
−
μ
i
1
′
)
(
1
−
μ
i
2
′
)
.
.
.
(
1
−
μ
i
n
′
)
,
ν
i
1
′
ν
i
2
′
.
.
.
ν
i
n
′
⟩
\tilde{d}_{i} = \left\langle\mu^{'}_{i},\nu^{'}_{i}\right\rangle = \sum_{j=1}^{n} \left\langle\mu^{'}_{ij},\nu^{'}_{ij}\right\rangle = \left\langle1-(1-\mu^{'}_{i1})(1-\mu^{'}_{i2})...(1-\mu^{'}_{in}),\nu^{'}_{i1}\nu^{'}_{i2}...\nu^{'}_{in}\right\rangle
d~i=⟨μi′,νi′⟩=∑j=1n⟨μij′,νij′⟩=⟨1−(1−μi1′)(1−μi2′)...(1−μin′),νi1′νi2′...νin′⟩计算方案
Y
i
Y_{i}
Yi的加权直觉模糊综合属性值
d
~
i
(
i
=
1
,
2
,
.
.
.
,
m
)
\tilde{d}_{i}\left(i=1,2,...,m\right)
d~i(i=1,2,...,m);
步骤S6 计算方案
Y
i
Y_{i}
Yi的加权直觉模糊综合属性值
d
~
i
\tilde{d}_{i}
d~i的得分值
s
(
d
~
i
)
s\left(\tilde{d}_{i}\right)
s(d~i)和精确值
h
(
d
~
i
)
h\left(\tilde{d}_{i}\right)
h(d~i),确定
d
~
i
(
i
=
1
,
2
,
.
.
.
,
m
)
\tilde{d}_{i}\left(i=1,2,...,m\right)
d~i(i=1,2,...,m)的不增排序顺序,并利用排序结果对方案
Y
=
{
Y
1
,
Y
2
,
.
.
.
,
Y
m
}
Y=\left\{Y_{1},Y_{2},...,Y_{m}\right\}
Y={Y1,Y2,...,Ym}进行优劣排序。
示 例 3.1 \color{red}{示例3.1} 示例3.1
考虑企业质量管理体系有效性评价问题。企业进行质量管理体系运行有效性评价的主要目的式发现质量管理体系运行管理体系运行过程中不完善或不适应环境变化的情况,提高组织的管理能力和经营业绩。通常从质量方针目标( G 1 G_{1} G1)、产品质量稳定性( G 2 G_{2} G2)、质量改进与创新( G 3 G_{3} G3)、资源管理情况( G 4 G_{4} G4)、财务运行情况( G 5 G_{5} G5)等五个方面进行评价。假设通过市场调研和专家咨询获得五家企业 Y i ( i = 1 , 2 , 3 , 4 , 5 ) Y_{i}\left(i=1,2,3,4,5\right) Yi(i=1,2,3,4,5)关于属性 G j ( i = 1 , 2 , 3 , 4 , 5 ) G_{j}\left(i=1,2,3,4,5\right) Gj(i=1,2,3,4,5)的直觉模糊评价结果如下表所示。
G 1 G_1 G1 | G 2 G_2 G2 | G 3 G_3 G3 | G 4 G_4 G4 | G 5 G_5 G5 | |
---|---|---|---|---|---|
Y 1 Y_1 Y1 | ⟨ 0.3 , 0.4 ⟩ \langle0.3,0.4\rangle ⟨0.3,0.4⟩ | ⟨ 0.2 , 0.2 ⟩ \langle0.2,0.2\rangle ⟨0.2,0.2⟩ | ⟨ 0.2 , 0.4 ⟩ \langle0.2,0.4\rangle ⟨0.2,0.4⟩ | ⟨ 0.3 , 0.5 ⟩ \langle0.3,0.5\rangle ⟨0.3,0.5⟩ | ⟨ 0.4 , 0.5 ⟩ \langle0.4,0.5\rangle ⟨0.4,0.5⟩ |
Y 2 Y_2 Y2 | ⟨ 0.4 , 0.2 ⟩ \langle0.4,0.2\rangle ⟨0.4,0.2⟩ | ⟨ 0.4 , 0.3 ⟩ \langle0.4,0.3\rangle ⟨0.4,0.3⟩ | ⟨ 0.3 , 0.4 ⟩ \langle0.3,0.4\rangle ⟨0.3,0.4⟩ | ⟨ 0.6 , 0.2 ⟩ \langle0.6,0.2\rangle ⟨0.6,0.2⟩ | ⟨ 0.8 , 0.1 ⟩ \langle0.8,0.1\rangle ⟨0.8,0.1⟩ |
Y 3 Y_3 Y3 | ⟨ 0.3 , 0.5 ⟩ \langle0.3,0.5\rangle ⟨0.3,0.5⟩ | ⟨ 0.5 , 0.2 ⟩ \langle0.5,0.2\rangle ⟨0.5,0.2⟩ | ⟨ 0.6 , 0.3 ⟩ \langle0.6,0.3\rangle ⟨0.6,0.3⟩ | ⟨ 0.5 , 0.2 ⟩ \langle0.5,0.2\rangle ⟨0.5,0.2⟩ | ⟨ 0.9 , 0.0 ⟩ \langle0.9,0.0\rangle ⟨0.9,0.0⟩ |
Y 4 Y_4 Y4 | ⟨ 0.6 , 0.3 ⟩ \langle0.6,0.3\rangle ⟨0.6,0.3⟩ | ⟨ 0.7 , 0.2 ⟩ \langle0.7,0.2\rangle ⟨0.7,0.2⟩ | ⟨ 0.4 , 0.4 ⟩ \langle0.4,0.4\rangle ⟨0.4,0.4⟩ | ⟨ 0.4 , 0.1 ⟩ \langle0.4,0.1\rangle ⟨0.4,0.1⟩ | ⟨ 0.7 , 0.2 ⟩ \langle0.7,0.2\rangle ⟨0.7,0.2⟩ |
Y 5 Y_5 Y5 | ⟨ 0.6 , 0.1 ⟩ \langle0.6,0.1\rangle ⟨0.6,0.1⟩ | ⟨ 0.3 , 0.1 ⟩ \langle0.3,0.1\rangle ⟨0.3,0.1⟩ | ⟨ 0.1 , 0.4 ⟩ \langle0.1,0.4\rangle ⟨0.1,0.4⟩ | ⟨ 0.7 , 0.1 ⟩ \langle0.7,0.1\rangle ⟨0.7,0.1⟩ | ⟨ 0.5 , 0.2 ⟩ \langle0.5,0.2\rangle ⟨0.5,0.2⟩ |
根据专家咨询、问卷调查等方式得到属性 G j ( i = 1 , 2 , 3 , 4 , 5 ) G_{j}\left(i=1,2,3,4,5\right) Gj(i=1,2,3,4,5)的直觉模糊权重向量为 ω ~ = ( ⟨ 0.35 , 0.45 ⟩ , ⟨ 0.65 , 0.20 ⟩ , ⟨ 0.45 , 0.25 ⟩ , ⟨ 0.40 , 0.30 ⟩ , ⟨ 0.25 , 0.45 ⟩ ) T \tilde{\omega}={\left(\left\langle0.35,0.45\right\rangle,\left\langle0.65,0.20\right\rangle,\left\langle0.45,0.25\right\rangle,\left\langle0.40,0.30\right\rangle,\left\langle0.25,0.45\right\rangle\right)}^{T} ω~=(⟨0.35,0.45⟩,⟨0.65,0.20⟩,⟨0.45,0.25⟩,⟨0.40,0.30⟩,⟨0.25,0.45⟩)T。那么企业 Y i ( i = 1 , 2 , 3 , 4 , 5 ) Y_{i}\left(i=1,2,3,4,5\right) Yi(i=1,2,3,4,5)的质量管理体系有效性的优劣排序结果如何?
代码如下:
import numpy as np
# 计算加权直觉模糊决策矩阵F'
def calculate_weight_IF_matrix(IFD_matrix,IFW_attr):
# IFD_matrix的shape为(m×2×n),分别表示m个方案、n个属性的直觉模糊数
# IFW_attr的shape为(2×n)表示n个属性的直觉模糊数
# 计算 [u_ij_'] = [p_j] * [u_ij]
# 计算 [v_ij_'] = [t_j] + [v_ij] - [t_j] * [v_ij]
weighted_IFD_matrix = np.zeros(IFD_matrix.shape)
for i in range(IFD_matrix.shape[0]):
weighted_IFD_matrix[i][0] = IFW_attr[0]*IFD_matrix[i][0]
weighted_IFD_matrix[i][1] = IFW_attr[1] + IFD_matrix[i][1] - IFW_attr[1]*IFD_matrix[i][1]
return weighted_IFD_matrix
# 计算综合评价结果
def calculate_assessment_result(weighted_IFD_matrix):
# weighted_IFD_matrix为加权直觉模糊决策矩阵,尺寸为(m×2×n),其中m表示方案数、n表示属性数
# 初始化结果矩阵,尺寸为(2×m)
result_matrix = np.zeros((2,weighted_IFD_matrix.shape[0]))
for i in range(weighted_IFD_matrix.shape[0]):
mu_i = weighted_IFD_matrix[i][0] # 第i个方案的mu值
nu_i = weighted_IFD_matrix[i][1] # 第i个方案的nu值
mu_r = 1
nu_r = 1
# 遍历mu_i,计算mu_tmp = (1-mu_1)*(1-mu_2)*...*(1-mu_n)
for value in mu_i:
mu_r *= 1-value
# 遍历nu_i,计算nu_result = nu_1*nu_2*...*nu_n
for value in nu_i:
nu_r *= value
# mu_result = 1 - mu_tmp
result_matrix[0][i] = 1 - mu_r
result_matrix[1][i] = nu_r
# 返回结果矩阵
return result_matrix
# 计算得分值并对方案进行排序
def calculate_scores_and_ranks(result_matrix):
# 计算得分值:
scores = result_matrix[0] - result_matrix[1]
# 获取排序索引
index = np.argsort(-scores)
return scores,index
# S1.直觉模糊决策矩阵F
IFD_matrix = np.array([[[0.3,0.2,0.2,0.3,0.4],[0.4,0.2,0.4,0.5,0.5]],
[[0.4,0.4,0.3,0.6,0.8],[0.2,0.3,0.4,0.2,0.1]],
[[0.3,0.5,0.6,0.5,0.9],[0.5,0.2,0.3,0.2,0.0]],
[[0.6,0.7,0.4,0.4,0.7],[0.3,0.2,0.4,0.1,0.2]],
[[0.6,0.3,0.1,0.7,0.5],[0.1,0.1,0.4,0.1,0.2]]])
print('直觉模糊决策矩阵F:\n',IFD_matrix)
# S2.各属性的直觉模糊权重向量
IFW_attr = np.array([[0.35,0.65,0.45,0.40,0.25],[0.45,0.20,0.25,0.30,0.45]])
print('\n各属性的直觉模糊权重向量:\n',IFW_attr)
# S3.计算加权直觉模糊决策矩阵F'
weighted_IFD_matrix = calculate_weight_IF_matrix(IFD_matrix,IFW_attr)
print('\n加权直觉模糊决策矩阵F:\n',weighted_IFD_matrix)
# S4.计算直觉模糊综合评价结果R
Result = calculate_assessment_result(weighted_IFD_matrix)
print('\直觉模糊综合评价结果R:\n',Result)
# S5.计算得分值并对方案进行排序
scores,index = calculate_scores_and_ranks(Result)
print('\n各方案得分值:\n',scores)
print('排序结果:\n',index+1)
计算结果如下:
直觉模糊决策矩阵F:
[[[0.3 0.2 0.2 0.3 0.4]
[0.4 0.2 0.4 0.5 0.5]]
[[0.4 0.4 0.3 0.6 0.8]
[0.2 0.3 0.4 0.2 0.1]]
[[0.3 0.5 0.6 0.5 0.9]
[0.5 0.2 0.3 0.2 0. ]]
[[0.6 0.7 0.4 0.4 0.7]
[0.3 0.2 0.4 0.1 0.2]]
[[0.6 0.3 0.1 0.7 0.5]
[0.1 0.1 0.4 0.1 0.2]]]
各属性的直觉模糊权重向量:
[[0.35 0.65 0.45 0.4 0.25]
[0.45 0.2 0.25 0.3 0.45]]
加权直觉模糊决策矩阵F:
[[[0.105 0.13 0.09 0.12 0.1 ]
[0.67 0.36 0.55 0.65 0.725]]
[[0.14 0.26 0.135 0.24 0.2 ]
[0.56 0.44 0.55 0.44 0.505]]
[[0.105 0.325 0.27 0.2 0.225]
[0.725 0.36 0.475 0.44 0.45 ]]
[[0.21 0.455 0.18 0.16 0.175]
[0.615 0.36 0.55 0.37 0.56 ]]
[[0.21 0.195 0.045 0.28 0.125]
[0.505 0.28 0.55 0.37 0.56 ]]]
\直觉模糊综合评价结果R:
[[0.43881137 0.66530451 0.72657302 0.75533566 0.61738068]
[0.06251603 0.03011254 0.02454705 0.02523074 0.01611394]]
各方案得分值:
[0.37629535 0.63519197 0.70202597 0.73010491 0.60126674]
排序结果:
[4 3 2 5 1]