4.1 问题描述
设某多属性决策问题有 m m m个方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m),组成方案集 Y = { Y 1 , Y 2 , ⋯ , Y m } Y=\left\{Y_{1},Y_{2},\cdots,Y_{m}\right\} Y={Y1,Y2,⋯,Ym}评价每个方案的属性(或指标)为 G j ( j = 1 , 2 , ⋯ , n ) G_{j}\left(j=1,2,\cdots,n\right) Gj(j=1,2,⋯,n),记属性集为 G = { G 1 , G 2 , ⋯ , G n } G=\left\{G_{1},G_{2},\cdots,G_{n}\right\} G={G1,G2,⋯,Gn},属性 G j ( j = 1 , 2 , ⋯ , n ) G_{j}\left(j=1,2,\cdots,n\right) Gj(j=1,2,⋯,n)的权重向量 ω = ( ω 1 , ω 2 , ⋯ , ω n ) \boldsymbol{\omega}=\left(\omega_{1},\omega_{2},\cdots,\omega_{n}\right) ω=(ω1,ω2,⋯,ωn)完全未知或信息不完全。假设方案 Y i ∈ Y Y_{i} \in Y Yi∈Y关于属性 G j ∈ G G_{j} \in G Gj∈G的评价值可以表示为 F ~ i j = ⟨ [ μ i j L , μ i j U ] , [ ν i j L , ν i j U ] ⟩ ( i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n ) \tilde{F}_{ij} = \left\langle \left[\mu_{ijL},\mu_{ijU}\right],\left[\nu_{ijL},\nu_{ijU}\right] \right\rangle \left(i=1,2,\cdots,m; j=1,2,\cdots,n\right) F~ij=⟨[μijL,μijU],[νijL,νijU]⟩(i=1,2,⋯,m;j=1,2,⋯,n), t i l d e F i j tilde{F}_{ij} tildeFij为直觉模糊集,其中 [ μ i j L , μ i j U ] \left[\mu_{ijL},\mu_{ijU}\right] [μijL,μijU]和 [ ν i j L , ν i j U ] \left[\nu_{ijL},\nu_{ijU}\right] [νijL,νijU]分别表示方案 Y i ∈ Y Y_{i} \in Y Yi∈Y满足属性 G j ∈ G G_{j} \in G Gj∈G和不满足属性 G j ∈ G G_{j} \in G Gj∈G的程度,且 0 ≤ μ i j + ν i j ≤ 1 0 \leq \mu_{ij}+\nu_{ij} \leq 1 0≤μij+νij≤1。则矩阵 F = ( ⟨ [ μ i j L , μ i j U ] , [ ν i j L , ν i j U ] ⟩ ) m × n F = \left( \left\langle \left[\mu_{ijL},\mu_{ijU}\right],\left[\nu_{ijL},\nu_{ijU}\right] \right\rangle \right)_{m×n} F=(⟨[μijL,μijU],[νijL,νijU]⟩)m×n为该多属性决策问题的直觉模糊决策矩阵。
现在假设决策者对方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)有一定的主观偏好,设主观偏好值为直觉模糊数 θ ~ i = ⟨ [ α i L , α i U ] , [ β i L , β i U ] ⟩ \tilde{\theta}_{i} = \left\langle \left[\alpha_{iL},\alpha_{iU}\right],\left[\beta_{iL},\beta_{iU}\right] \right\rangle θ~i=⟨[αiL,αiU],[βiL,βiU]⟩,现在的问题是依据直觉模糊决策矩阵 F F F,如何通过确定属性权重 ω = ( ω 1 , ω 2 , ⋯ , ω n ) T \boldsymbol{\omega} = {\left(\omega_{1},\omega_{2},\cdots,\omega_{n}\right)}^{T} ω=(ω1,ω2,⋯,ωn)T,得到一个有效的决策分析方法对所有方案进行优劣排序。
4.2 属性权重的确定方法
由于种种条件的制约,决策者的主观偏好与客观偏好之间往往存在一定的差距。为了使决策更具合理性,属性权重的选择应使决策者的主观偏好值与客观偏好值(属性值)的总偏差最小化。
对于属性 G i ( j = 1 , 2 , ⋯ , n ) G_{i}\left(j=1,2,\cdots,n\right) Gi(j=1,2,⋯,n),如果 d ( F ~ i j , θ ~ i ) d\left(\tilde{F}_{ij},\tilde{\theta}_{i}\right) d(F~ij,θ~i)表示决策者对方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)的主观偏好值 θ ~ i \tilde{\theta}_{i} θ~i与相应的客观偏好值(属性值) F ~ i j \tilde{F}_{ij} F~ij)之间的偏差,可定义为:
d ( F ~ i j , θ ~ i ) = 1 4 ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ( i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n ) (4.1) d\left(\tilde{F}_{ij},\tilde{\theta}_{i}\right) = \frac{1}{4}\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right) \left(i=1,2,\cdots,m;j=1,2,\cdots,n\right) \tag{4.1} d(F~ij,θ~i)=41ωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(i=1,2,⋯,m;j=1,2,⋯,n)(4.1)
则对属性 G i ( j = 1 , 2 , ⋯ , n ) G_{i}\left(j=1,2,\cdots,n\right) Gi(j=1,2,⋯,n)而言,所有方案的主观偏好值与客观偏好值(属性值)之间的偏差为:
D j ( ω ) = ∑ i = 1 m d ( F ~ i j , θ ~ i ) = 1 4 ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ( i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n ) (4.2) \begin{aligned} D_{j}\left(\omega\right) &= \sum_{i=1}^{m} d\left(\tilde{F}_{ij},\tilde{\theta}_{i}\right) \\ &= \frac{1}{4}\sum_{i=1}^{m} {\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)} \left(i=1,2,\cdots,m;j=1,2,\cdots,n\right) \tag{4.2} \end{aligned} Dj(ω)=i=1∑md(F~ij,θ~i)=41i=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(i=1,2,⋯,m;j=1,2,⋯,n)(4.2)
那么,所有方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)对所有属性 G j ( j = 1 , 2 , ⋯ , n ) G_{j}\left(j=1,2,\cdots,n\right) Gj(j=1,2,⋯,n)的主观偏好值与客观偏好值(属性值)之间的总偏差为
D ( ω ) = ∑ j = 1 n D j ( ω ) = 1 4 ∑ j = 1 n ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ( i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n ) (4.3) \color{blue} { \begin{aligned} D\left(\omega\right) &= \sum_{j=1}^{n} D_{j}\left(\omega\right) \\ &= \frac{1}{4} \sum_{j=1}^{n} \sum_{i=1}^{m} {\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)} \left(i=1,2,\cdots,m;j=1,2,\cdots,n\right) \tag{4.3} \end{aligned} } D(ω)=j=1∑nDj(ω)=41j=1∑ni=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(i=1,2,⋯,m;j=1,2,⋯,n)(4.3)
由此,权重向量 ω \omega ω的选择应使 D ( ω ) D\left(\omega\right) D(ω)最小化,即
min D ( ω ) = 1 4 ∑ j = 1 n ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ( i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n ) (4.4) \color{blue} { \min D\left(\omega\right) = \frac{1}{4}\sum_{j=1}^{n} \sum_{i=1}^{m} {\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)} \left(i=1,2,\cdots,m;j=1,2,\cdots,n\right) \tag{4.4} } minD(ω)=41j=1∑ni=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(i=1,2,⋯,m;j=1,2,⋯,n)(4.4)
当 属 性 权 重 完 全 未 知 \color{red}{属性权重完全未知} 属性权重完全未知时,可以建立如下最优化模型
{ min D ( ω ) = 1 4 ∑ j = 1 n ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) s.t. ∑ j = 1 n ω j 2 = 1 , ω j ≥ 0 ( j = 1 , 2 , ⋯ , n ) (4.5) \color{blue} { \left\{ \begin{aligned} & \min D\left(\omega\right) = \frac{1}{4}\sum_{j=1}^{n} \sum_{i=1}^{m} {\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)} \\ & \text {s.t.} \sum_{j=1}^{n}\omega_{j}^{2} = 1, \omega_{j} \geq 0 \left(j=1,2,\cdots,n\right) \end{aligned} \right. \tag{4.5} } ⎩⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎧minD(ω)=41j=1∑ni=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)s.t.j=1∑nωj2=1,ωj≥0(j=1,2,⋯,n)(4.5)
为了求解最优化模型 ( 4.5 ) \left(4.5\right) (4.5),构造拉格朗日函数
L ( ω , λ ) = 1 4 ∑ j = 1 n ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) + λ 8 ( ∑ j = 1 n ω j 2 − 1 ) (4.6) L\left(\omega,\lambda\right)=\frac{1}{4} {\sum_{j=1}^{n} \sum_{i=1}^{m} {\omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)}}+\frac{\lambda}{8}\left(\sum_{j=1}^{n} {\omega_{j}^{2}-1}\right) \tag{4.6} L(ω,λ)=41j=1∑ni=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)+8λ(j=1∑nωj2−1)(4.6)
对其求偏导数,并令偏导数等于0,可得
{ ∂ L ∂ ω j = 1 4 ∑ i = 1 m ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) + 1 4 λ ω j = 0 ∂ L ∂ λ = 1 4 ( ∑ j = 1 n ω j 2 − 1 ) = 0 (4.7) \color{blue} { \left\{ \begin{aligned} & \frac{\partial{L}}{\partial{\omega_{j}}} = \frac{1}{4} \sum_{i=1}^{m}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)+\frac{1}{4} \lambda \omega_{j}=0 \\ & \frac{\partial{L}}{\partial{\lambda}}=\frac{1}{4}\left(\sum_{j=1}^{n} \omega_{j}^{2}-1\right)=0 \end{aligned} \right. \tag{4.7} } ⎩⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎧∂ωj∂L=41i=1∑m(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)+41λωj=0∂λ∂L=41(j=1∑nωj2−1)=0(4.7)
则有
ω j ∗ = ∑ i = 1 m ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ∑ j = 1 n [ ∑ i = 1 m ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ] 2 (4.8) \color{blue} { \omega_{j}^{*}=\frac{\sum_{i=1}^{m}{\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)}}{\sqrt{\sum_{j=1}^{n}\left[\sum_{i=1}^{m}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)\right]^{2}}} \tag{4.8} } ωj∗=∑j=1n[∑i=1m(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)]2∑i=1m(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(4.8)
对 ω j ∗ \omega_{j}^{*} ωj∗进行归一化处理,可以得到属性权重:
ω j = ∑ i = 1 m ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) ∑ j = 1 n ∑ i = 1 m ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) (4.9) \color{red} { \omega_{j}=\frac{\sum_{i=1}^{m}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)}{\sum_{j=1}^{n} \sum_{i=1}^{m}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right)} \tag{4.9} } ωj=∑j=1n∑i=1m(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)∑i=1m(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)(4.9)
&ems;当属性权重信息不完全确定时,如果权重向量 ω = ( ω 1 , ω 2 , ⋯ , ω n ) T \boldsymbol{\omega} = \left(\omega_{1},\omega_{2},\cdots,\omega_{n}\right)^{T} ω=(ω1,ω2,⋯,ωn)T满足条件 ω ‾ j ≤ ω j ω ‾ j , ∑ j = 1 n ω j = 1 , ω j ≥ 0 ( j = 1 , 2 , ⋯ , n ) \underline{\omega}_{j} \leq \omega_{j} \overline{\omega}_{j}, \sum_{j=1}^{n} \omega_{j} = 1, \omega_{j} \geq 0 \left(j=1,2,\cdots,n\right) ωj≤ωjωj,∑j=1nωj=1,ωj≥0(j=1,2,⋯,n),则求解权重向量 ω \omega ω等价于求解以下最优化模型:
{ max D ( ω ) = ∑ j = 1 n D j ( ω ) = 1 4 ∑ j = 1 n ∑ i = 1 m ω j ( ∣ μ i j L − α i L ∣ + ∣ μ i j U − α i U ∣ + ∣ ν i j L − β i L ∣ + ∣ ν i j U − β i U ∣ ) s.t. ω ‾ j ⩽ ω j ≤ ω ˉ j , ∑ j = 1 n ω j = 1 , ω j ≥ 0 ( j = 1 , 2 , ⋯ , n ) (4.10) \color{red} { \left\{ \begin{aligned} & \max D(\omega)=\sum_{j=1}^{n} D_{j}(\omega)=\frac{1}{4} \sum_{j=1}^{n} \sum_{i=1}^{m} \omega_{j}\left(|\mu_{ijL}-\alpha_{iL}| + |\mu_{ijU}-\alpha_{iU}| + |\nu_{ijL}-\beta_{iL}| + |\nu_{ijU}-\beta_{iU}|\right) \\ & \text {s.t.} \underline{\omega}_{j} \leqslant \omega_{j} \leq \bar{\omega}_{j}, \sum_{j=1}^{n} \omega_{j}=1, \omega_{j} \geq 0 \left(j=1,2, \cdots, n\right) \end{aligned} \right. \tag{4.10} } ⎩⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎧maxD(ω)=j=1∑nDj(ω)=41j=1∑ni=1∑mωj(∣μijL−αiL∣+∣μijU−αiU∣+∣νijL−βiL∣+∣νijU−βiU∣)s.t.ωj⩽ωj≤ωˉj,j=1∑nωj=1,ωj≥0(j=1,2,⋯,n)(4.10)
4.3 分析步骤
根据以上分析,决策者对方案有主观偏好情形下的直觉模糊多属性决策步骤可归纳如下:
S.1 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},\cdots, Y_{m}\right\} Y={Y1,Y2,⋯,Ym}和属性集 G = { G 1 , G 2 , ⋯ , G n } G = \left\{G_{1},G_{2},\cdots, G_{n}\right\} G={G1,G2,⋯,Gn}。
S.2 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,同时决策者给出对决策方案 Y i ∈ Y Y_{i} \in Y Yi∈Y的主观偏好值 θ ~ i = ⟨ [ α i L , α i U ] , [ β i L , β i U ] ⟩ \tilde{\theta}_{i} = \left\langle\left[\alpha_{iL},\alpha_{iU}\right],\left[\beta_{iL},\beta_{iU}\right]\right\rangle θ~i=⟨[αiL,αiU],[βiL,βiU]⟩。
S.3 利用式 ( 4.9 ) \left(4.9\right) (4.9)或求解最优化模型 ( 4.10 ) \left(4.10\right) (4.10)确定各属性的权重,得到属性权重向量 ω = ( ω 1 , ω 2 , ⋯ , ω n ) T \omega = {\left(\omega_{1},\omega_{2},\cdots,\omega_{n}\right)}^{T} ω=(ω1,ω2,⋯,ωn)T。
S.4 计算方案 Y i Y_{i} Yi的综合属性值 d ~ i \tilde{d}_{i} d~i:
d ~ i = ⟨ μ i , ν i ⟩ = IIFHA ω ( F ~ i 1 , F ~ i 2 , ⋯ , F ~ i n ) = ⟨ [ 1 − ∏ j = 1 n ( 1 − μ i j L ) ω j , 1 − ∏ j = 1 n ( 1 − μ i j U ) ω j ] , [ ∏ j = 1 n ( ν i j L ) ω j , ∏ j = 1 n ( ν i j U ) ω j ] ⟩ (3.44) \color{red} { \begin{aligned} \tilde{d}_{i} &= \left\langle \mu_{i},\nu_{i} \right\rangle \\ &= \operatorname{IIFHA}_{\omega}\left(\tilde{F}_{i1},\tilde{F}_{i2},\cdots,\tilde{F}_{in}\right) \\ &= \left\langle \left[ 1-\prod_{j=1}^{n}{\left(1-\mu_{ijL}\right)}^{\omega_{j}},1-\prod_{j=1}^{n}{\left(1-\mu_{ijU}\right)}^{\omega_{j}} \right], \left[\prod_{j=1}^{n}{\left(\nu_{ijL}\right)}^{\omega_{j}},\prod_{j=1}^{n}{\left(\nu_{ijU}\right)}^{\omega_{j}} \right] \right\rangle \tag{3.44} \end{aligned} } d~i=⟨μi,νi⟩=IIFHAω(F~i1,F~i2,⋯,F~in)=⟨[1−j=1∏n(1−μijL)ωj,1−j=1∏n(1−μijU)ωj],[j=1∏n(νijL)ωj,j=1∏n(νijU)ωj]⟩(3.44)
或
d ~ i = ⟨ μ i , ν i ⟩ = IIFHG ω ( F ~ i 1 , F ~ i 2 , ⋯ , F ~ i n ) = ⟨ [ ∏ j = 1 n ( μ i j L ) ω j , ∏ j = 1 n ( μ i j U ) ω j ] , [ 1 − ∏ j = 1 n ( 1 − ν i j L ) ω j , 1 − ∏ j = 1 n ( 1 − ν i j U ) ω j ] ⟩ (3.45) \color{red} { \begin{aligned} \tilde{d}_{i} &= \left\langle \mu_{i},\nu_{i} \right\rangle \\ &= \operatorname{IIFHG}_{\omega}\left(\tilde{F}_{i1},\tilde{F}_{i2},\cdots,\tilde{F}_{in}\right) \\ &= \left\langle \left[ \prod_{j=1}^{n}{\left(\mu_{ijL}\right)}^{\omega_{j}},\prod_{j=1}^{n}{\left(\mu_{ijU}\right)}^{\omega_{j}} \right], \left[1 - \prod_{j=1}^{n}{\left(1 - \nu_{ijL}\right)}^{\omega_{j}},1 - \prod_{j=1}^{n}{\left(1 - \nu_{ijU}\right)}^{\omega_{j}} \right] \right\rangle \tag{3.45} \end{aligned} } d~i=⟨μi,νi⟩=IIFHGω(F~i1,F~i2,⋯,F~in)=⟨[j=1∏n(μijL)ωj,j=1∏n(μijU)ωj],[1−j=1∏n(1−νijL)ωj,1−j=1∏n(1−νijU)ωj]⟩(3.45)
S.5 利用直觉模糊数的得分函数和精确度公式,计算方案
Y
i
Y_{i}
Yi的综合属性值
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,\cdots,m}\right)
d~i(i=1,2,⋯,m)的不增排列顺序,并利用排序结果对方案
Y
i
(
i
=
1
,
2
,
⋯
,
m
)
Y_{i}\left(i=1,2,\cdots,m\right)
Yi(i=1,2,⋯,m)进行优劣排序。
s
(
d
~
i
)
=
μ
i
L
+
μ
i
U
−
ν
i
L
−
ν
i
U
2
h
(
d
~
i
)
=
μ
i
L
+
μ
i
U
+
ν
i
L
+
ν
i
U
2
\begin{aligned} & \text{s}\left(\tilde{d}_{i}\right) = \frac{\mu_{iL}+\mu_{iU}-\nu_{iL}-\nu_{iU}}{2}\\ & \text{h}\left(\tilde{d}_{i}\right) = \frac{\mu_{iL}+\mu_{iU}+\nu_{iL}+\nu_{iU}}{2} \end{aligned}
s(d~i)=2μiL+μiU−νiL−νiUh(d~i)=2μiL+μiU+νiL+νiU
4.4 实例分析
考虑突发事件应急预案评估问题。突发事件应急预案是针对各种突发事件类型而事先制订的一套能迅速、有效、有序地解决问题的行动计划或方案,为全面、客观地评判应急预案处置突发事件的能力,应从预案处置的快速性( G 1 G_1 G1)、预案内容的合理性( G 2 G_2 G2)、预案保障的充分性( G 3 G_3 G3)、预案消耗费用的合理性( G 4 G_4 G4)以及预案的广泛适用性( G 5 G_5 G5)等五个方面进行综合评价。假设现有五个应急预案 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),专家组根据自己的知识、经验以及已有的统计数据确定出每个应急预案 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 ( j = 1 , 2 , 3 , 4.5 ) G_{j}\left(j=1,2,3,4.5\right) Gj(j=1,2,3,4.5)的区间直觉模糊评价信息,得到的区间直觉模糊决策矩阵 F = ( ⟨ [ μ i j L , μ i j U ] , [ ν i j L , ν i j U ] ⟩ ) 5 × 5 F=\left(\left\langle\left[\mu_{ijL},\mu_{ijU}\right],\left[\nu_{ijL},\nu_{ijU}\right]\right\rangle\right)_{5×5} F=(⟨[μijL,μijU],[νijL,νijU]⟩)5×5 如表3-6 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 | ⟨ [ 0.5 , 0.6 ] , [ 0.1 , 0.3 ] ⟩ \left\langle[0.5,0.6],[0.1,0.3]\right\rangle ⟨[0.5,0.6],[0.1,0.3]⟩ | ⟨ [ 0.3 , 0.4 ] , [ 0.4 , 0.5 ] ⟩ \left\langle[0.3,0.4],[0.4,0.5]\right\rangle ⟨[0.3,0.4],[0.4,0.5]⟩ | ⟨ [ 0.1 , 0.3 ] , [ 0.5 , 0.6 ] ⟩ \left\langle[0.1,0.3],[0.5,0.6]\right\rangle ⟨[0.1,0.3],[0.5,0.6]⟩ | ⟨ [ 0.7 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \left\langle[0.7,0.8],[0.1,0.2]\right\rangle ⟨[0.7,0.8],[0.1,0.2]⟩ | ⟨ [ 0.5 , 0.7 ] , [ 0.1 , 0.3 ] ⟩ \left\langle[0.5,0.7],[0.1,0.3]\right\rangle ⟨[0.5,0.7],[0.1,0.3]⟩ |
Y 2 Y_2 Y2 | ⟨ [ 0.4 , 0.5 ] , [ 0.2 , 0.4 ] ⟩ \left\langle[0.4,0.5],[0.2,0.4]\right\rangle ⟨[0.4,0.5],[0.2,0.4]⟩ | ⟨ [ 0.4 , 0.6 ] , [ 0.2 , 0.4 ] ⟩ \left\langle[0.4,0.6],[0.2,0.4]\right\rangle ⟨[0.4,0.6],[0.2,0.4]⟩ | ⟨ [ 0.3 , 0.5 ] , [ 0.4 , 0.5 ] ⟩ \left\langle[0.3,0.5],[0.4,0.5]\right\rangle ⟨[0.3,0.5],[0.4,0.5]⟩ | ⟨ [ 0.6 , 0.7 ] , [ 0.2 , 0.3 ] ⟩ \left\langle[0.6,0.7],[0.2,0.3]\right\rangle ⟨[0.6,0.7],[0.2,0.3]⟩ | ⟨ [ 0.7 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \left\langle[0.7,0.8],[0.1,0.2]\right\rangle ⟨[0.7,0.8],[0.1,0.2]⟩ |
Y 3 Y_3 Y3 | ⟨ [ 0.6 , 0.6 ] , [ 0.2 , 0.3 ] ⟩ \left\langle[0.6,0.6],[0.2,0.3]\right\rangle ⟨[0.6,0.6],[0.2,0.3]⟩ | ⟨ [ 0.7 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \left\langle[0.7,0.8],[0.1,0.2]\right\rangle ⟨[0.7,0.8],[0.1,0.2]⟩ | ⟨ [ 0.4 , 0.5 ] , [ 0.3 , 0.5 ] ⟩ \left\langle[0.4,0.5],[0.3,0.5]\right\rangle ⟨[0.4,0.5],[0.3,0.5]⟩ | ⟨ [ 0.4 , 0.6 ] , [ 0.1 , 0.3 ] ⟩ \left\langle[0.4,0.6],[0.1,0.3]\right\rangle ⟨[0.4,0.6],[0.1,0.3]⟩ | ⟨ [ 0.6 , 0.7 ] , [ 0.2 , 0.3 ] ⟩ \left\langle[0.6,0.7],[0.2,0.3]\right\rangle ⟨[0.6,0.7],[0.2,0.3]⟩ |
Y 4 Y_4 Y4 | ⟨ [ 0.3 , 0.5 ] , [ 0.3 , 0.4 ] ⟩ \left\langle[0.3,0.5],[0.3,0.4]\right\rangle ⟨[0.3,0.5],[0.3,0.4]⟩ | ⟨ [ 0.5 , 0.7 ] , [ 0.2 , 0.3 ] ⟩ \left\langle[0.5,0.7],[0.2,0.3]\right\rangle ⟨[0.5,0.7],[0.2,0.3]⟩ | ⟨ [ 0.6 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \left\langle[0.6,0.8],[0.1,0.2]\right\rangle ⟨[0.6,0.8],[0.1,0.2]⟩ | ⟨ [ 0.5 , 0.6 ] , [ 0.2 , 0.3 ] ⟩ \left\langle[0.5,0.6],[0.2,0.3]\right\rangle ⟨[0.5,0.6],[0.2,0.3]⟩ | ⟨ [ 0.7 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \left\langle[0.7,0.8],[0.1,0.2]\right\rangle ⟨[0.7,0.8],[0.1,0.2]⟩ |
Y 5 Y_5 Y5 | ⟨ [ 0.3 , 0.4 ] , [ 0.4 , 0.5 ] ⟩ \left\langle[0.3,0.4],[0.4,0.5]\right\rangle ⟨[0.3,0.4],[0.4,0.5]⟩ | ⟨ [ 0.2 , 0.4 ] , [ 0.3 , 0.4 ] ⟩ \left\langle[0.2,0.4],[0.3,0.4]\right\rangle ⟨[0.2,0.4],[0.3,0.4]⟩ | ⟨ [ 0.3 , 0.4 ] , [ 0.3 , 0.4 ] ⟩ \left\langle[0.3,0.4],[0.3,0.4]\right\rangle ⟨[0.3,0.4],[0.3,0.4]⟩ | ⟨ [ 0.3 , 0.5 ] , [ 0.4 , 0.5 ] ⟩ \left\langle[0.3,0.5],[0.4,0.5]\right\rangle ⟨[0.3,0.5],[0.4,0.5]⟩ | ⟨ [ 0.4 , 0.6 ] , [ 0.2 , 03 ] ⟩ \left\langle[0.4,0.6],[0.2,03]\right\rangle ⟨[0.4,0.6],[0.2,03]⟩ |
如果决策者对方案 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)的主观偏好值为直觉模糊数,分别为 θ ~ 1 = ⟨ [ 0.3 , 0.4 ] , [ 0.4 , 0.5 ] ⟩ \tilde{\theta}_{1}=\left\langle[0.3,0.4],[0.4,0.5]\right\rangle θ~1=⟨[0.3,0.4],[0.4,0.5]⟩, θ ~ 1 = ⟨ [ 0.5 , 0.6 ] , [ 0.2 , 0.3 ] ⟩ \tilde{\theta}_{1}=\left\langle[0.5,0.6],[0.2,0.3]\right\rangle θ~1=⟨[0.5,0.6],[0.2,0.3]⟩, θ ~ 1 = ⟨ [ 0.4 , 0.5 ] , [ 0.2 , 0.4 ] ⟩ \tilde{\theta}_{1}=\left\langle[0.4,0.5],[0.2,0.4]\right\rangle θ~1=⟨[0.4,0.5],[0.2,0.4]⟩, θ ~ 1 = ⟨ [ 0.7 , 0.8 ] , [ 0.1 , 0.2 ] ⟩ \tilde{\theta}_{1}=\left\langle[0.7,0.8],[0.1,0.2]\right\rangle θ~1=⟨[0.7,0.8],[0.1,0.2]⟩, θ ~ 1 = ⟨ [ 0.4 , 0.6 ] , [ 0.2 , 0.3 ] ⟩ \tilde{\theta}_{1}=\left\langle[0.4,0.6],[0.2,0.3]\right\rangle θ~1=⟨[0.4,0.6],[0.2,0.3]⟩。下面对企业 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进行优劣排序。
代码如下:
## 考虑专家主观偏好的区间直觉模糊评价--<权重信息完全未知>
import numpy as np
# 属性权重计算函数
def calculate_attribute_weight(IIFDM,ORIIFM):
'''
输入参数:
IFDM:觉模糊决策矩阵,尺寸为(m,n,4),其中m表示方案数,n表示属性数
ORIFM:决策者对各个方案的主观偏好直觉模糊矩阵,尺寸为(m,4),其中m表示方案数
返回值:
attribute_weight:属性权重向量,尺寸为(n)
'''
# 按公式3.30计算各个属性的权重
attribute_weight = np.zeros((IIFDM.shape[1]))
sum_all = 0
for j in range(IIFDM.shape[1]):
sum_j = 0
for i in range(IIFDM.shape[0]):
sum_j += np.abs(IIFDM[i][j][0] - ORIIFM[i][0]) + np.abs(IIFDM[i][j][1] - ORIIFM[i][1]) + \
np.abs(IIFDM[i][j][2] - ORIIFM[i][2]) + np.abs(IIFDM[i][j][3] - ORIIFM[i][3])
attribute_weight[j] = sum_j
sum_all += sum_j
attribute_weight /= sum_all
# 返回结果
return attribute_weight
# 各方案的综合评价直觉模糊数计算函数
def calculate_comprehensive_IIFM(IIFDM,attribute_weight,operator_type='IIFWA'):
'''
输入参数:
IIFDM:觉模糊决策矩阵,尺寸为(m,n,4),其中m表示方案数,n表示属性数
attribute_weight:属性权重向量,尺寸为(n)
operator_type:操作算子类型,可取值“IIFWA”或“IIFWG”,分别表示直
觉模糊加权平均算子、直觉模糊加权几何算子,默认值为
“IIFWA”
返回值:
comprehensive_IIFM:综合评价直觉模糊矩阵,尺寸为(m,4)
'''
# 初始化
comprehensive_IIFM = np.zeros((IIFDM.shape[0],4))
# 计算综合评价直觉模糊矩阵
if operator_type == 'IIFWA': # IIFWA算子
for i in range(IIFDM.shape[0]):
mu_L = 1
mu_U = 1
nu_L = 1
nu_U = 1
for j in range(IIFDM.shape[1]):
mu_L *= np.power(1-IIFDM[i][j][0],attribute_weight[j])
mu_U *= np.power(1-IIFDM[i][j][1],attribute_weight[j])
nu_L *= np.power(IIFDM[i][j][2],attribute_weight[j])
nu_U *= np.power(IIFDM[i][j][3],attribute_weight[j])
comprehensive_IIFM[i][0] = 1 - mu_L
comprehensive_IIFM[i][1] = 1 - mu_U
comprehensive_IIFM[i][2] = nu_L
comprehensive_IIFM[i][3] = nu_U
elif operator_type == 'IIFWG': # IIFWG算子
for i in range(IIFDM.shape[0]):
mu_L = 1
mu_U = 1
nu_L = 1
nu_U = 1
for j in range(IIFDM.shape[1]):
mu_L *= np.power(IIFDM[i][j][0],attribute_weight[j])
mu_U *= np.power(IIFDM[i][j][1],attribute_weight[j])
nu_L *= np.power(1 - IIFDM[i][j][2],attribute_weight[j])
nu_U *= np.power(1 - IIFDM[i][j][3],attribute_weight[j])
comprehensive_IIFM[i][0] = mu_L
comprehensive_IIFM[i][1] = mu_U
comprehensive_IIFM[i][2] = 1 - nu_L
comprehensive_IIFM[i][3] = 1 - nu_U
else:
print("输入操作算子类型错误:operator_type参数可取值“IIFWA”或“IIFWG”!")
# 返回计算结果
return comprehensive_IIFM
# 方案得分值与排序函数
def calculate_score_and_sort(comprehensive_IIFM):
'''
输入参数:
comprehensive_IIFM:综合评价直觉模糊矩阵,尺寸为(m,2),m表示方案数
返回值:
scores:各方案的得分值,尺寸为(1,m)
sorted_index:降序排序结果,尺寸为(1,m)
'''
# 计算得分值
scores = np.zeros((comprehensive_IIFM.shape[0]))
for i in range(comprehensive_IIFM.shape[0]):
scores[i] = (comprehensive_IIFM[i][0] + comprehensive_IIFM.shape[1] - comprehensive_IIFM[i][2] - \
comprehensive_IIFM[i][3])/2
# 降序排序
sorted_index = np.argsort(-scores)
# 返回计算结果
return scores,sorted_index
# S.0: 确定直觉模糊评价决策矩阵IFDM和专家对各个方案的主观偏好直觉模糊矩阵ORIFM
IIFDM = np.array([[[0.5,0.6,0.1,0.3],[0.3,0.4,0.4,0.5],[0.1,0.3,0.5,0.6],[0.7,0.8,0.1,0.2],[0.5,0.7,0.1,0.3]],
[[0.4,0.5,0.2,0.4],[0.4,0.6,0.2,0.4],[0.3,0.5,0.4,0.5],[0.6,0.7,0.2,0.3],[0.7,0.8,0.1,0.2]],
[[0.6,0.6,0.2,0.3],[0.7,0.8,0.1,0.2],[0.4,0.5,0.3,0.5],[0.4,0.6,0.1,0.3],[0.6,0.7,0.2,0.3]],
[[0.3,0.5,0.3,0.4],[0.5,0.7,0.2,0.3],[0.6,0.8,0.1,0.2],[0.5,0.6,0.2,0.3],[0.7,0.8,0.1,0.2]],
[[0.3,0.4,0.4,0.5],[0.2,0.4,0.3,0.4],[0.3,0.4,0.3,0.4],[0.3,0.5,0.4,0.5],[0.4,0.6,0.2,0.3]]])
ORIIFM = np.array([[0.3,0.4,0.4,0.5],
[0.5,0.6,0.2,0.3],
[0.4,0.5,0.2,0.4],
[0.7,0.8,0.1,0.2],
[0.4,0.6,0.2,0.3]])
# S.1:计算各属性的权重
attribute_weight = calculate_attribute_weight(IIFDM,ORIIFM)
print("\n","#"*75)
print("各属性的权重向量为:\n",attribute_weight)
print("#"*75)
# S.2:计算各方案的综合属性直觉模糊数
comprehensive_IIFM_IIFWA = calculate_comprehensive_IIFM(IIFDM,attribute_weight)
comprehensive_IIFM_IIFWG = calculate_comprehensive_IIFM(IIFDM,attribute_weight,operator_type='IIFWG')
print("\n","#"*75)
print("各方案的综合评价直觉模糊矩阵(IIFWA)为:\n",comprehensive_IIFM_IIFWA)
print("各方案的综合评价直觉模糊矩阵(IIFWG)为:\n",comprehensive_IIFM_IIFWG)
print("#"*75)
# S.3:计算各方案的得分值并按降序排序
scores_IIFWA,sorted_index_IIFWA = calculate_score_and_sort(comprehensive_IIFM_IIFWA)
scores_IIFWG,sorted_index_IIFWG = calculate_score_and_sort(comprehensive_IIFM_IIFWG)
print("\n","#"*75)
print("各方案的综合得分值(IIFWA)为:\n",scores_IIFWA)
print("各方案的综合排序(IIFWA)为:\n\n",sorted_index_IIFWA + 1)
print("各方案的综合得分值(IIFWG)为:\n",scores_IIFWG)
print("各方案的综合排序(IIFWG)为:\n",sorted_index_IIFWG + 1)
print("#"*75)
计算结果如下:
###########################################################################
各属性的权重向量为:
[0.265625 0.171875 0.15625 0.2421875 0.1640625]
###########################################################################
###########################################################################
各方案的综合评价直觉模糊矩阵(IIFWA)为:
[[0.48685146 0.62251743 0.16318984 0.33085701]
[0.50274764 0.63415462 0.19891988 0.34479094]
[0.55254691 0.64927648 0.15991862 0.30305359]
[0.51443652 0.67647772 0.17839413 0.28438316]
[0.30163583 0.46286362 0.32484541 0.42734079]]
各方案的综合评价直觉模糊矩阵(IIFWG)为:
[[0.38638127 0.552194 0.23424385 0.37475171]
[0.46244368 0.60458602 0.22024078 0.36541381]
[0.52419922 0.62840044 0.17736139 0.32043203]
[0.47466546 0.64364241 0.19819148 0.29872098]
[0.29332787 0.45125327 0.33836888 0.43904828]]
###########################################################################
###########################################################################
各方案的综合得分值(IIFWA)为:
[1.9964023 1.97951841 2.04478735 2.02582961 1.77472482]
各方案的综合排序(IIFWA)为:
[3 4 1 2 5]
各方案的综合得分值(IIFWG)为:
[1.88869286 1.93839455 2.0132029 1.9888765 1.75795536]
各方案的综合排序(IIFWG)为:
[3 4 2 1 5]
###########################################################################