4.1 动态直觉模糊多属性决策TOPSIS方法
4.1.1 动态直觉模糊集成算子
设 A ~ ( t k ) = ⟨ μ k , ν k ⟩ ( k = 1 , 2 , ⋯ , p ) \tilde{A}\left(t_{k}\right) = \left\langle \mu_{k},\nu_{k} \right\rangle \left(k=1,2,\cdots,p\right) A~(tk)=⟨μk,νk⟩(k=1,2,⋯,p)是 p p p个不同时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的动态直觉模糊数, w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)为时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重向量,满足 ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p \sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p ∑k=1pwk=1,wk≥0,k=1,2,⋯,p,则称
DIFWA ω , w ( A ~ ( t 1 ) , A ~ ( t 2 ) , ⋯ , A ~ ( t p ) ) = ⟨ 1 − ∏ k = 1 p ( 1 − μ k ) w k , ∏ k = 1 p ( ν k ) w k ⟩ (4.1) \color{blue} { \operatorname{DIFWA}_{\omega, w}\left(\tilde{A}\left(t_{1}\right), \tilde{A}\left(t_{2}\right), \cdots, \tilde{A}\left(t_{p}\right)\right)=\left\langle 1-\prod_{k=1}^{p}\left(1-\mu_{k}\right)^{w_{k}}, \prod_{k=1}^{p}\left(\nu_{k}\right)^{w_{k}}\right\rangle \tag{4.1} } DIFWAω,w(A~(t1),A~(t2),⋯,A~(tp))=⟨1−k=1∏p(1−μk)wk,k=1∏p(νk)wk⟩(4.1)
为动态直觉模糊加权平均算子。
设 A ~ ( t k ) = ⟨ μ k , ν k ⟩ ( k = 1 , 2 , ⋯ , p ) \tilde{A}\left(t_{k}\right) = \left\langle \mu_{k},\nu_{k} \right\rangle \left(k=1,2,\cdots,p\right) A~(tk)=⟨μk,νk⟩(k=1,2,⋯,p)是 p p p个不同时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的动态直觉模糊数, w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)为时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重向量,满足 ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p \sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p ∑k=1pwk=1,wk≥0,k=1,2,⋯,p,则称
DIFWG ω , w ( A ~ ( t 1 ) , A ~ ( t 2 ) , ⋯ , A ~ ( t p ) ) = ⟨ ∏ k = 1 p ( μ k ) w k , 1 − ∏ k = 1 p ( 1 − ν k ) w k ⟩ (4.2) \color{blue} { \operatorname{DIFWG}_{\omega, w}\left(\tilde{A}\left(t_{1}\right), \tilde{A}\left(t_{2}\right), \cdots, \tilde{A}\left(t_{p}\right)\right)=\left\langle \prod_{k=1}^{p}\left(\mu_{k}\right)^{w_{k}}, 1-\prod_{k=1}^{p}\left(1-\nu_{k}\right)^{w_{k}}\right\rangle \tag{4.2} } DIFWGω,w(A~(t1),A~(t2),⋯,A~(tp))=⟨k=1∏p(μk)wk,1−k=1∏p(1−νk)wk⟩(4.2)
为动态直觉模糊加权几何算子。
4.1.2 属性权重与时段权重的确定方法
属性的权重可采用直觉模糊集的模糊熵进行计算,时段权重可采用基本单位区间单调函数法、正态分布法、指数分布法、平均年龄法以及基于“厚今薄古”法。其中“厚今薄古”法确定时段 t k ( k = 1 , 2 , ⋯ , p ) t_{k}\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重,即求解非线性规划模型:
{ max I = − ∑ k = 1 p w k log w k s.t. λ = ∑ k = 1 p p − k p − 1 , ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p (4.3) \color{blue} { \left\{ \begin{aligned} & \max{I} = - \sum_{k=1}^{p}w_{k}\log{w_{k}} \\ & \text{s.t.} \lambda = \sum_{k=1}^{p}\frac{p-k}{p-1},\sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p \end{aligned} \right. \tag{4.3} } ⎩⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎧maxI=−k=1∑pwklogwks.t.λ=k=1∑pp−1p−k,k=1∑pwk=1,wk≥0,k=1,2,⋯,p(4.3)
式中, λ \lambda λ为时间度,表示决策者对时段的重视程度,通常用 λ = 0.1 , 0.2 , 0.3 , 0.4 , 0.5 \lambda=0.1,0.2,0.3,0.4,0.5 λ=0.1,0.2,0.3,0.4,0.5代表决策者极端重视、强烈重视、明显重视、稍微重视近期数据和同样重视所有时段数据。
4.1.3 动态直觉模糊多属性决策步骤
S.1 确定动态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 m ) G=\left(G_{1},G_{2},\cdots,G_{m}\right) G=(G1,G2,⋯,Gm),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,构建动态多属性决策问题各时段 t k t_{k} tk的直觉模糊决策矩阵 F ( t k ) F\left(t_{k}\right) F(tk)
S.2 给定时间度 λ \lambda λ,求解非线性规划模型 ( 4.3 ) (4.3) (4.3),得到时段 t k t_{k} tk的权重 w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)。
S.3 利用动态直觉模糊加权平均算子( DIFWA \operatorname{DIFWA} DIFWA),计算方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)在各时段 t k ( k = 1 , 2 , ⋯ , p ) t_{k}\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)直觉模糊属性值 F ~ i j ( t k ) \tilde{F}_{ij}\left(t_{k}\right) F~ij(tk)的综合值 F ~ i j \tilde{F}_{ij} F~ij:
F ~ i j = ⟨ μ i j , ν i j ⟩ = DIFWA ω , w ( F ~ i j ( t 1 ) , F ~ i j ( t 2 ) , ⋯ , F ~ i j ( t p ) ) = ⟨ 1 − ∏ k = 1 p ( 1 − μ i j ( k ) ) w t , ∏ k = 1 p ( ν i j ( k ) ) w t j ⟩ , i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n (4.4) \color{red} { \begin{aligned} \tilde{F}_{ij} &= \left\langle \mu_{ij}, \nu_{ij} \right\rangle \\ &= \operatorname{DIFWA}_{\omega, w}\left(\tilde{F}_{ij}\left(t_{1}\right), \tilde{F}_{ij}\left(t_{2}\right), \cdots, \tilde{F}_{ij}\left(t_{p}\right)\right) \\ &= \left\langle { 1-\prod_{k=1}^{p}\left(1-\mu_{ij}^{(k)}\right)^{w_{t}}, \prod_{k=1}^{p}\left(\nu_{ij}^{(k)}\right)^{w_{tj}} }\right\rangle, i=1,2, \cdots, m; j=1,2, \cdots, n \\ \tag{4.4} \end{aligned} } F~ij=⟨μij,νij⟩=DIFWAω,w(F~ij(t1),F~ij(t2),⋯,F~ij(tp))=⟨1−k=1∏p(1−μij(k))wt,k=1∏p(νij(k))wtj⟩,i=1,2,⋯,m;j=1,2,⋯,n(4.4)
得到直觉模糊综合决策矩阵 F = ( ⟨ μ i j , ν i j ⟩ ) m × n F=\left(\left\langle\mu_{ij},\nu_{ij}\right\rangle\right)_{m×n} F=(⟨μij,νij⟩)m×n
S.4 根据直觉模糊综合决策矩阵 F F F,确定动态多属性决策问题的正理想解 Y + Y^{+} Y+和负理想解 Y − Y^{-} Y−:
Y + = ( ⟨ μ 1 + , ν 1 + ⟩ , ⟨ μ 2 + , ν 2 + ⟩ , ⋯ , ⟨ μ n + , ν n + ⟩ ) = ( ⟨ max i μ i 1 , min i ν i n ⟩ , ⟨ max i μ i 2 , min i ν 12 ⟩ , ⋯ , ⟨ max i μ i n , min i ν i n ⟩ ) (4.5) \begin{aligned} Y^{+} &=\left(\left\langle\mu_{1}^{+}, \nu_{1}^{+}\right\rangle,\left\langle\mu_{2}^{+}, \nu_{2}^{+}\right\rangle, \cdots,\left\langle\mu_{n}^{+}, \nu_{n}^{+}\right\rangle\right) \\ &=\left(\left\langle\max_{i} \mu_{i 1}, \min_{i} \nu_{i n}\right\rangle,\left\langle\max_{i} \mu_{i 2}, \min_{i} \nu_{12}\right\rangle, \cdots,\left\langle\max_{i} \mu_{i n}, \min_{i} \nu_{i n}\right\rangle\right) \tag{4.5} \end{aligned} Y+=(⟨μ1+,ν1+⟩,⟨μ2+,ν2+⟩,⋯,⟨μn+,νn+⟩)=(⟨imaxμi1,iminνin⟩,⟨imaxμi2,iminν12⟩,⋯,⟨imaxμin,iminνin⟩)(4.5)
Y − = ( ⟨ μ 1 − , ν 1 − ⟩ , ⟨ μ 2 − , ν 2 − ⟩ , ⋯ , ⟨ μ n − , ν n − ⟩ ) = ( ⟨ min i μ i 1 , max i ν i n ⟩ , ⟨ min i μ i 2 , max i ν 12 ⟩ , ⋯ , ⟨ min i μ i n , max i ν i n ⟩ ) (4.6) \begin{aligned} Y^{-} &=\left(\left\langle\mu_{1}^{-}, \nu_{1}^{-}\right\rangle,\left\langle\mu_{2}^{-}, \nu_{2}^{-}\right\rangle, \cdots,\left\langle\mu_{n}^{-}, \nu_{n}^{-}\right\rangle\right) \\ &=\left(\left\langle\min_{i} \mu_{i 1}, \max_{i} \nu_{i n}\right\rangle,\left\langle\min_{i} \mu_{i 2}, \max_{i} \nu_{12}\right\rangle, \cdots,\left\langle\min_{i} \mu_{i n}, \max_{i} \nu_{i n}\right\rangle\right) \tag{4.6} \end{aligned} Y−=(⟨μ1−,ν1−⟩,⟨μ2−,ν2−⟩,⋯,⟨μn−,νn−⟩)=(⟨iminμi1,imaxνin⟩,⟨iminμi2,imaxν12⟩,⋯,⟨iminμin,imaxνin⟩)(4.6)
S.5 利用式 ( 2.3 ) (2.3) (2.3)、式 ( 2.4 ) (2.4) (2.4)计算属性 G j ( j = 1 , 2 , ⋯ , n ) G_{j}\left(j=1,2,\cdots,n\right) Gj(j=1,2,⋯,n)的模糊熵 E j E_{j} Ej和权重 ω j \omega_{j} ωj:
E j = 1 m ∑ i = 1 m cos π ( μ i j − ν i j ) ( 1 − π i j ) 2 , j = 1 , 2 , ⋯ , n (4.7) E_{j}=\frac{1}{m} \sum_{i=1}^{m} \cos \frac{\pi\left(\mu_{i j}-\nu_{i j}\right)\left(1-\pi_{i j}\right)}{2}, \quad j=1,2, \cdots, n \tag{4.7} Ej=m1i=1∑mcos2π(μij−νij)(1−πij),j=1,2,⋯,n(4.7)
ω j = 1 − E j n − ∑ j = 1 n E j , j = 1 , 2 , ⋯ , n (4.8) \omega_{j}=\frac{1-E_{j}}{n-\sum_{j=1}^{n} E_{j}}, \quad j=1,2, \cdots, n \tag{4.8} ωj=n−∑j=1nEj1−Ej,j=1,2,⋯,n(4.8)
S.6 利用直觉模糊综合决策矩阵 F F F,计算各方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)到正理想解 Y + Y^{+} Y+和负理想解 Y − Y^{-} Y−的距离 d i + d_{i}^{+} di+和 d i − d_{i}^{-} di−:
d i + = 1 2 ∑ j = 1 n ω j [ ∣ μ i j − μ j + ∣ + ∣ ν i j − v j ∗ ∣ + ∣ π i − π j + ∣ ] (4.9) d_{i}^{+}=\frac{1}{2} \sum_{j=1}^{n} \omega_{j}\left[\left|\mu_{i j}-\mu_{j}^{+}\right|+\left|\nu_{i j}-v_{j}^{*}\right|+\left|\pi_{i}-\pi_{j}^{+}\right|\right] \tag{4.9} di+=21j=1∑nωj[∣∣μij−μj+∣∣+∣∣νij−vj∗∣∣+∣∣πi−πj+∣∣](4.9)
d i − = 1 2 ∑ j = 1 n ω j [ ∣ μ i j − μ j − ∣ + ∣ ν i j − v j − ∣ + ∣ π i j − π j − ∣ ] t a g 4.10 d_{i}^{-}=\frac{1}{2} \sum_{j=1}^{n} \omega_{j}\left[\left|\mu_{i j}-\mu_{j}^{-}\right|+\left|\nu_{i j}-v_{j}^{-}\right|+\left|\pi_{i j}-\pi_{j}^{-}\right|\right] tag{4.10} di−=21j=1∑nωj[∣∣μij−μj−∣∣+∣∣νij−vj−∣∣+∣∣πij−πj−∣∣]tag4.10
S.7 计算方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)的贴近度 c i c_{i} ci:
c i = d i − d i − + d i + (4.11) c_{i} = \frac {d_{i}^{-}} {d_{i}^{-} + d_{i}^{+}} \tag{4.11} ci=di−+di+di−(4.11)
并利用贴近度c的大小对方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)进行排序, c i c_i ci越大表明方案 Y i Y_{i} Yi离正理想解越近、离负理想解越远,方案越优。
4.2 动态区间直觉模糊多属性决策TOPSIS方法
4.2.1 动态区间直觉模糊集成算子**
设 A ~ ( t k ) = ⟨ [ μ k L , μ k U ] , [ ν k L , ν k U ] ⟩ ( k = 1 , 2 , ⋯ , p ) \tilde{A}\left(t_{k}\right) = \left\langle \left[\mu_{kL},\mu_{kU}\right],\left[\nu_{kL},\nu_{kU}\right] \right\rangle \left(k=1,2,\cdots,p\right) A~(tk)=⟨[μkL,μkU],[νkL,νkU]⟩(k=1,2,⋯,p)是 p p p个不同时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的动态区间直觉模糊数, w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)为时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重向量,满足 ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p \sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p ∑k=1pwk=1,wk≥0,k=1,2,⋯,p,则称
DIIFWA ω , w ( A ~ ( t 1 ) , A ~ ( t 2 ) , ⋯ , A ~ ( t p ) ) = ⟨ [ 1 − ∏ k = 1 p ( 1 − μ k L ) w k , 1 − ∏ k = 1 p ( 1 − μ k U ) w k ] , [ ∏ k = 1 p ( ν k L ) w k , ∏ k = 1 p ( ν k U ) w k ] ⟩ (4.12) \color{blue} { \operatorname{DIIFWA}_{\omega, w}\left(\tilde{A}\left(t_{1}\right), \tilde{A}\left(t_{2}\right), \cdots, \tilde{A}\left(t_{p}\right)\right)=\left\langle \left[1-\prod_{k=1}^{p}\left(1-\mu_{kL}\right)^{w_{k}},1-\prod_{k=1}^{p}\left(1-\mu_{kU}\right)^{w_{k}}\right], \left[\prod_{k=1}^{p}\left(\nu_{kL}\right)^{w_{k}},\prod_{k=1}^{p}\left(\nu_{kU}\right)^{w_{k}}\right] \right\rangle \tag{4.12} } DIIFWAω,w(A~(t1),A~(t2),⋯,A~(tp))=⟨[1−k=1∏p(1−μkL)wk,1−k=1∏p(1−μkU)wk],[k=1∏p(νkL)wk,k=1∏p(νkU)wk]⟩(4.12)
为动态区间直觉模糊加权平均算子。
设 A ~ ( t k ) = ⟨ [ μ k L , μ k U ] , [ ν k L , ν k U ] ⟩ ( k = 1 , 2 , ⋯ , p ) \tilde{A}\left(t_{k}\right) = \left\langle \left[\mu_{kL},\mu_{kU}\right],\left[\nu_{kL},\nu_{kU}\right] \right\rangle \left(k=1,2,\cdots,p\right) A~(tk)=⟨[μkL,μkU],[νkL,νkU]⟩(k=1,2,⋯,p)是 p p p个不同时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的动态区间直觉模糊数, w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)为时段 t k ( k = 1 , 2 , ⋯ , p ) t_k\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重向量,满足 ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p \sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p ∑k=1pwk=1,wk≥0,k=1,2,⋯,p,则称
DIIFWG ω , w ( A ~ ( t 1 ) , A ~ ( t 2 ) , ⋯ , A ~ ( t p ) ) = ⟨ [ ∏ k = 1 p ( μ k L ) w k , ∏ k = 1 p ( μ k U ) w k ] , [ 1 − ∏ k = 1 p ( 1 − ν k L ) w k , 1 − ∏ k = 1 p ( 1 − ν k U ) w k ] ⟩ (4.13) \color{blue} { \operatorname{DIIFWG}_{\omega, w}\left(\tilde{A}\left(t_{1}\right), \tilde{A}\left(t_{2}\right), \cdots, \tilde{A}\left(t_{p}\right)\right)=\left\langle \left[\prod_{k=1}^{p}\left(\mu_{kL}\right)^{w_{k}},\prod_{k=1}^{p}\left(\mu_{kU}\right)^{w_{k}}\right], \left[1-\prod_{k=1}^{p}\left(1-\nu_{kL}\right)^{w_{k}},1-\prod_{k=1}^{p}\left(1-\nu_{kU}\right)^{w_{k}}\right] \right\rangle \tag{4.13} } DIIFWGω,w(A~(t1),A~(t2),⋯,A~(tp))=⟨[k=1∏p(μkL)wk,k=1∏p(μkU)wk],[1−k=1∏p(1−νkL)wk,1−k=1∏p(1−νkU)wk]⟩(4.13)
为动态区间直觉模糊加权平均算子。
4.2.2 属性权重与时段权重的确定方法
属性的权重可采用区间直觉模糊集的模糊熵进行计算,时段权重可采用基本单位区间单调函数法、正态分布法、指数分布法、平均年龄法以及基于“厚今薄古”法。其中“厚今薄古”法确定时段 t k ( k = 1 , 2 , ⋯ , p ) t_{k}\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)的权重,即求解非线性规划模型:
{ max I = − ∑ k = 1 p w k log w k s.t. λ = ∑ k = 1 p p − k p − 1 , ∑ k = 1 p w k = 1 , w k ≥ 0 , k = 1 , 2 , ⋯ , p (4.14) \color{blue} { \left\{ \begin{aligned} & \max{I} = - \sum_{k=1}^{p}w_{k}\log{w_{k}} \\ & \text{s.t.} \lambda = \sum_{k=1}^{p}\frac{p-k}{p-1},\sum_{k=1}^{p}w_{k}=1,w_{k}\geq 0, k=1,2,\cdots,p \end{aligned} \right. \tag{4.14} } ⎩⎪⎪⎪⎪⎪⎨⎪⎪⎪⎪⎪⎧maxI=−k=1∑pwklogwks.t.λ=k=1∑pp−1p−k,k=1∑pwk=1,wk≥0,k=1,2,⋯,p(4.14)
式中, λ \lambda λ为时间度,表示决策者对时段的重视程度,通常用 λ = 0.1 , 0.2 , 0.3 , 0.4 , 0.5 \lambda=0.1,0.2,0.3,0.4,0.5 λ=0.1,0.2,0.3,0.4,0.5代表决策者极端重视、强烈重视、明显重视、稍微重视近期数据和同样重视所有时段数据。
4.2.3 动态区间直觉模糊多属性决策步骤
S.1 确定动态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 m ) G=\left(G_{1},G_{2},\cdots,G_{m}\right) G=(G1,G2,⋯,Gm),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的区间直觉模糊特征信息,构建动态多属性决策问题各时段 t k t_{k} tk的区间直觉模糊决策矩阵 F ( t k ) F\left(t_{k}\right) F(tk)
S.2 给定时间度 λ \lambda λ,求解非线性规划模型 ( 4.3 ) (4.3) (4.3),得到时段 t k t_{k} tk的权重 w k ( k = 1 , 2 , ⋯ , p ) w_{k}\left(k=1,2,\cdots,p\right) wk(k=1,2,⋯,p)。
S.3 利用动态区间直觉模糊加权平均算子( DIIFWA \operatorname{DIIFWA} DIIFWA),计算方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)在各时段 t k ( k = 1 , 2 , ⋯ , p ) t_{k}\left(k=1,2,\cdots,p\right) tk(k=1,2,⋯,p)区间直觉模糊属性值 F ~ i j ( t k ) \tilde{F}_{ij}\left(t_{k}\right) F~ij(tk)的综合值 F ~ i j \tilde{F}_{ij} F~ij:
F ~ i j = ⟨ [ μ i j L , μ i j U ] , [ ν i j L , ν i j U ] ⟩ = DIIFWA ω , w ( F ~ i j ( t 1 ) , F ~ i j ( t 2 ) , ⋯ , F ~ i j ( t p ) ) = ⟨ [ 1 − ∏ k = 1 p ( 1 − μ i j L ( k ) ) w t , 1 − ∏ k = 1 p ( 1 − μ i j U ( k ) ) w t ] , [ ∏ k = 1 p ( ν i j L ( k ) ) w t j , ∏ k = 1 p ( ν i j L ( k ) ) w t j ] ⟩ , i = 1 , 2 , ⋯ , m ; j = 1 , 2 , ⋯ , n (4.15) \color{red} { \begin{aligned} \tilde{F}_{ij} &= \left\langle \left[ \mu_{ijL},\mu_{ijU} \right], \left[ \nu_{ijL},\nu_{ijU} \right] \right\rangle \\ &= \operatorname{DIIFWA}_{\omega, w}\left(\tilde{F}_{ij}\left(t_{1}\right), \tilde{F}_{ij}\left(t_{2}\right), \cdots, \tilde{F}_{ij}\left(t_{p}\right)\right) \\ &= \left\langle \left[ 1-\prod_{k=1}^{p}\left(1-\mu_{ijL}^{(k)}\right)^{w_{t}}, 1-\prod_{k=1}^{p}\left(1-\mu_{ijU}^{(k)}\right)^{w_{t}} \right], \left[\prod_{k=1}^{p}\left(\nu_{ijL}^{(k)}\right)^{w_{tj}},\prod_{k=1}^{p}\left(\nu_{ijL}^{(k)}\right)^{w_{tj}}\right] \right\rangle, \\ &i=1,2, \cdots, m; j=1,2, \cdots, n \\ \tag{4.15} \end{aligned} } F~ij=⟨[μijL,μijU],[νijL,νijU]⟩=DIIFWAω,w(F~ij(t1),F~ij(t2),⋯,F~ij(tp))=⟨[1−k=1∏p(1−μijL(k))wt,1−k=1∏p(1−μijU(k))wt],[k=1∏p(νijL(k))wtj,k=1∏p(νijL(k))wtj]⟩,i=1,2,⋯,m;j=1,2,⋯,n(4.15)
得到区间直觉模糊综合决策矩阵 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
S.4 根据区间直觉模糊综合决策矩阵 F F F,确定动态多属性决策问题的正理想解 Y + Y^{+} Y+和负理想解 Y − Y^{-} Y−:
Y + = ( ⟨ [ μ 1 L + , μ 1 U + ] , [ ν 1 L + , ν 1 U + ] ⟩ , ⟨ [ μ 2 L + , μ 2 U + ] , [ ν 2 L + , ν 2 U + ] ⟩ , ⋯ , ⟨ [ μ n L + , μ n U + ] , [ ν n L + , ν n U + ] ⟩ ) (4.16) \begin{aligned} Y^{+} &=\left(\left\langle \left[ \mu_{1L}^{+},\mu_{1U}^{+}\right],\left[\nu_{1L}^{+},\nu_{1U}^{+}\right] \right\rangle,\left\langle \left[ \mu_{2L}^{+},\mu_{2U}^{+}\right],\left[\nu_{2L}^{+},\nu_{2U}^{+}\right] \right\rangle, \cdots,\left\langle \left[ \mu_{nL}^{+},\mu_{nU}^{+}\right],\left[\nu_{nL}^{+},\nu_{nU}^{+}\right] \right\rangle\right) \\ \tag{4.16} \end{aligned} Y+=(⟨[μ1L+,μ1U+],[ν1L+,ν1U+]⟩,⟨[μ2L+,μ2U+],[ν2L+,ν2U+]⟩,⋯,⟨[μnL+,μnU+],[νnL+,νnU+]⟩)(4.16)
Y − = ( ⟨ [ μ 1 L − , μ 1 U − ] , [ ν 1 L − , ν 1 U − ] ⟩ , ⟨ [ μ 2 L − , μ 2 U − ] , [ ν 2 L − , ν 2 U − ] ⟩ , ⋯ , ⟨ [ μ n L + , μ n U − ] , [ ν n L + , ν n U − ] ⟩ ) (4.17) \begin{aligned} Y^{-} &=\left(\left\langle \left[ \mu_{1L}^{-},\mu_{1U}^{-}\right],\left[\nu_{1L}^{-},\nu_{1U}^{-}\right] \right\rangle,\left\langle \left[ \mu_{2L}^{-},\mu_{2U}^{-}\right],\left[\nu_{2L}^{-},\nu_{2U}^{-}\right] \right\rangle, \cdots,\left\langle \left[ \mu_{nL}^{+},\mu_{nU}^{-}\right],\left[\nu_{nL}^{+},\nu_{nU}^{-}\right] \right\rangle\right) \\ \tag{4.17} \end{aligned} Y−=(⟨[μ1L−,μ1U−],[ν1L−,ν1U−]⟩,⟨[μ2L−,μ2U−],[ν2L−,ν2U−]⟩,⋯,⟨[μnL+,μnU−],[νnL+,νnU−]⟩)(4.17)
式中
⟨ [ μ j L + , μ j U + ] , [ ν j L + , ν j U + ] ⟩ = ⟨ [ max j μ i j L , max j μ i j U ] , [ min j ν i j L , min j ν i j U ] ⟩ \left\langle \left[ \mu_{jL}^{+},\mu_{jU}^{+}\right],\left[\nu_{jL}^{+},\nu_{jU}^{+}\right] \right\rangle = \left\langle \left[ \max_{j}\mu_{ijL},\max_{j}\mu_{ijU}\right],\left[\min_{j}\nu_{ijL},\min_{j}\nu_{ijU}\right] \right\rangle ⟨[μjL+,μjU+],[νjL+,νjU+]⟩=⟨[jmaxμijL,jmaxμijU],[jminνijL,jminνijU]⟩
⟨ [ μ j L − , μ j U − ] , [ ν j L − , ν j U − ] ⟩ = ⟨ [ min j μ i j L , min j μ i j U ] , [ max j ν i j L , max j ν i j U ] ⟩ \left\langle \left[ \mu_{jL}^{-},\mu_{jU}^{-}\right],\left[\nu_{jL}^{-},\nu_{jU}^{-}\right] \right\rangle = \left\langle \left[ \min_{j}\mu_{ijL},\min_{j}\mu_{ijU}\right],\left[\max_{j}\nu_{ijL},\max_{j}\nu_{ijU}\right] \right\rangle ⟨[μjL−,μjU−],[νjL−,νjU−]⟩=⟨[jminμijL,jminμijU],[jmaxνijL,jmaxνijU]⟩
S.5 利用式 ( 3.3 ) (3.3) (3.3)、式 ( 3.4 ) (3.4) (3.4)计算属性 G j ( j = 1 , 2 , ⋯ , n ) G_{j}\left(j=1,2,\cdots,n\right) Gj(j=1,2,⋯,n)的模糊熵 E j E_{j} Ej和权重 ω j \omega_{j} ωj:
E j = 1 m ∑ i = 1 m cos π ( ∣ ( μ i j L − ν i j L ) ( 1 − π i j L ) ∣ + ∣ ( μ i j U − ν i j U ) ( 1 − π i j U ) ∣ ) 2 , j = 1 , 2 , ⋯ , n (4.18) E_{j}=\frac{1}{m} \sum_{i=1}^{m} \cos \frac{\pi\left( \left|\left(\mu_{ijL}-\nu_{ijL}\right)\left(1-\pi_{ijL}\right)\right| + \left|\left(\mu_{ijU}-\nu_{ijU}\right)\left(1-\pi_{ijU}\right)\right| \right)}{2}, \quad j=1,2, \cdots, n \tag{4.18} Ej=m1i=1∑mcos2π(∣(μijL−νijL)(1−πijL)∣+∣(μijU−νijU)(1−πijU)∣),j=1,2,⋯,n(4.18)
ω j = 1 − E j n − ∑ j = 1 n E j , j = 1 , 2 , ⋯ , n (4.19) \omega_{j}=\frac{1-E_{j}}{n-\sum_{j=1}^{n} E_{j}}, \quad j=1,2, \cdots, n \tag{4.19} ωj=n−∑j=1nEj1−Ej,j=1,2,⋯,n(4.19)
S.6 利用直觉模糊综合决策矩阵 F F F,计算各方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)到正理想解 Y + Y^{+} Y+和负理想解 Y − Y^{-} Y−的距离 d i + d_{i}^{+} di+和 d i − d_{i}^{-} di−:
d i + ( ω ) = 1 4 ∑ j = 1 n ω j [ ∣ μ i j L − μ j L + ∣ + ∣ μ i j U − μ j U + ∣ + ∣ ν i j L − ν j L + ∣ + ∣ ν i j U − ν j U + ∣ ] (4.20) d_{i}^{+}\left(\omega\right)=\frac{1}{4} \sum_{j=1}^{n} \omega_{j}\left[ \left|\mu_{ijL}-\mu_{jL}^{+}\right| + \left|\mu_{ijU}-\mu_{jU}^{+}\right| + \left|\nu_{ijL}-\nu_{jL}^{+}\right| + \left|\nu_{ijU}-\nu_{jU}^{+}\right| \right] \tag{4.20} di+(ω)=41j=1∑nωj[∣∣∣μijL−μjL+∣∣∣+∣∣∣μijU−μjU+∣∣∣+∣∣∣νijL−νjL+∣∣∣+∣∣∣νijU−νjU+∣∣∣](4.20)
d i − ( ω ) = 1 4 ∑ j = 1 n ω j [ ∣ μ i j L − μ j L − ∣ + ∣ μ i j U − μ j U − ∣ + ∣ ν i j L − ν j L − ∣ + ∣ ν i j U − ν j U − ∣ ] (4.21) d_{i}^{-}\left(\omega\right)=\frac{1}{4} \sum_{j=1}^{n} \omega_{j}\left[ \left|\mu_{ijL}-\mu_{jL}^{-}\right| + \left|\mu_{ijU}-\mu_{jU}^{-}\right| + \left|\nu_{ijL}-\nu_{jL}^{-}\right| + \left|\nu_{ijU}-\nu_{jU}^{-}\right| \right] \tag{4.21} di−(ω)=41j=1∑nωj[∣∣∣μijL−μjL−∣∣∣+∣∣∣μijU−μjU−∣∣∣+∣∣∣νijL−νjL−∣∣∣+∣∣∣νijU−νjU−∣∣∣](4.21)
S.7 计算方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)的贴近度 c i c_{i} ci:
c i = d i − d i − + d i + (4.22) c_{i} = \frac {d_{i}^{-}} {d_{i}^{-} + d_{i}^{+}} \tag{4.22} ci=di−+di+di−(4.22)
并利用贴近度c的大小对方案 Y i ( i = 1 , 2 , ⋯ , m ) Y_{i}\left(i=1,2,\cdots,m\right) Yi(i=1,2,⋯,m)进行排序, c i c_{i} ci越大表明方案 Y i Y_{i} Yi离正理想解越近、离负理想解越远,方案越优。