﻿ 船舶绿色智能技术应用综合评价方法
 舰船科学技术  2022, Vol. 44 Issue (23): 43-48    DOI: 10.3404/j.issn.1672-7649.2022.23.009 PDF

Research on the comprehensive evaluation method of green and intelligent technology application in ships
WU Ya-nan, WANG Li-zheng, WANG Min
School of Naval Architecture Ocean and Energy Power Engineering, Wuhan University of Technology, Wuhan 430063, China
Abstract: Aiming at the problem that there are many evaluation indexes of green intelligent technology application in ships and some index attribute values are fuzzy and stochastic, an uncertain multi-attribute index comprehensive evaluation method based on mixed stochastic dominance is proposed. In this method, the single index attribute value with fuzziness is determined by using the index attribute value quantification method combined with fuzzy linguistic variable-UOWA operator, and the single index attribute value with randomness is determined by the index attribute value quantification method of random simulation; on this basis, the mixed multi-attribute decision model is constructed, and the integrated mixed dominance matrix is established by determining the mixed dominance relationship and mixed dominance degree of each index, and The Promethee II method was used to select the optimal solution. Finally, the method was applied to 7500 t bulk carrier demonstration vessel in the Yangtze River to verify the effectiveness of the comprehensive evaluation method.
Key words: green intelligent ships     index quantification methodology     hybrid dominance     PROMETHEE II     comprehensive evaluation
0 引　言

1 船舶绿色智能技术应用综合评价思路

 图 1 船舶绿色智能技术应用综合评价指标 Fig. 1 Comprehensive evaluation index of green and intelligent technology application on ships

2 综合评价关键指标属性值量化方法

2.1 定性指标量化 2.1.1 语言评价集量化

2.1.2 模糊语言变量信息集结

 $p(a > b) = \frac{{{\text{min}}\left\{ {{a^U} - {a^L} + {b^U} - {b^L},{\text{max}}({a^U} - {b^L},0)} \right\}}}{{({a^U} - {a^L} + {b^U} - {b^L})}}\text{，}$ (1)

$a > b$ 的可能度。

 ${\text{UOWA}}\left( {{{\tilde x}_1},{{\tilde x}_2}, \cdots {{\tilde x}_n}} \right) = \sum\limits_{i = 1}^n {{v_i}{{\tilde y}_i}} \text{，}$ (2)

1）假设有s位专家参加评价，得到不同专家对不同指标的评价矩阵为

 ${{\boldsymbol{x}}_{kj}} = \left( {\begin{array}{*{20}{c}} {{x_{11}}}& \ldots &{{{\text{x}}_{1j}}} \\ \vdots & \ddots & \vdots \\ {{x_{s1}}}& \cdots &{{x_{sj}}} \end{array}} \right)\text{，}$ (3)

2）由下列公式确定加权向量

 $\begin{split} & v = {\left[ {{v_1},{v_2}, \cdots {v_i}} \right]^{\rm{T}}}\\ & {v_i} = Q\left( {\frac{i}{m}} \right) - Q\left( {\frac{{i - 1}}{n}} \right)\text{，} \end{split}$ (4)

 $Q\left( r \right) = \left\{ {\begin{array}{*{20}{c}} {\begin{array}{*{20}{c}} 0\text{，}&{r < a} \text{，} \end{array}} \\ {\begin{array}{*{20}{c}} {\dfrac{{r - a}}{{b - a}}}\text{，}&{a \leqslant r \leqslant b}\text{，} \end{array}} \\ {\begin{array}{*{20}{c}} 1\text{，}&{r > b}\text{。} \end{array}} \end{array}} \right.$ (5)

3）对所有专家对同一指标的评价值进行两两比较，获得可能度矩阵 ${\boldsymbol{P}} = {\left( {{p_{ij}}} \right)_{s \times s}}$ ，利用式(6)计算不同指标可能度矩阵的排序向量 $V = {\left[ {{v_1},{v_2}, \cdots {v_i}} \right]^{\rm{T}}}$ ，并按 ${v_i}$ 的大小对区间数 ${\tilde x_i}$ 排序得到 ${\tilde y_i}$

 ${v_i} = \frac{1}{{m\left( {m - 1} \right)}}\left( {\sum\limits_{j = 1}^m {{p_{ij}} + \frac{m}{2} - 1} } \right)\text{。}$ (6)

4）利用UOWA算子计算各指标的属性值评价结果。

2.2 定量指标量化

 图 2 经济性随机模拟计算流程图 Fig. 2 Flow chart of economic stochastic simulation calculation
3 不确定性混合多属性决策模型

 $C{\text{ = }}\sum\limits_{j = 1}^n {{w_j}{D_j}}\text{。}$ (7)

3.1 属性值归一化方法及累计分布函数

1）绿色性指标

 $x_{ij}' = \frac{{{x_{ij}} - \mathop {\min }\limits_{1 \leqslant i \leqslant m} {x_{ij}}}}{{\mathop {\max }\limits_{1 \leqslant i \leqslant m} {x_{ij}} - \mathop {\min }\limits_{1 \leqslant i \leqslant m} {x_{ij}}}} \text{，}$ (8)

 ${F_{ij}}\left( x \right) = \left\{ {\begin{array}{*{20}{c}} {\begin{array}{*{20}{c}} 0\text{，}&{x < x_{ij}'} \text{，} \end{array}} \\ {\begin{array}{*{20}{c}} 1\text{，}&{x \geqslant x_{ij}'} \text{。} \end{array}} \end{array}} \right.$ (9)

2）智能性指标

3）技术性指标

 ${F_{ij}}\left( x \right) = \int_{x_{ij}^L}^{x_{ij}^U} {{f_{ij}}} \left( x \right){\rm{d}}x \text{。}$ (10)

4）社会效益性指标

5）经济性指标

 \begin{aligned}[b] & {F_{ij}}\left( {{x'}} \right) = \int_0^{{x'}} {\left( {\mathop {\max }\limits_{1 \leqslant i \leqslant m} {x_{ij}} - \mathop {\min }\limits_{1 \leqslant i \leqslant m} {x_{ij}}} \right){f_{ij}}}\text{，} \\ & \left( {\mathop {\max }\limits_{1 \leqslant i \leqslant m} {x_{ij}} - \left( {\mathop {\max }\limits_{1 \leqslant i \leqslant m} {x_{ij}} - \mathop {\min }\limits_{1 \leqslant i \leqslant m} {x_{ij}}} \right){x'}} \right){\rm{d}}x \text{。} \end{aligned} (11)
3.2 混合占优度计算

 ${d}_{ik}^{j} = \left\{\begin{array}{llllllllllll} \frac{{\text{e}}^{{\displaystyle {\int }_{a}^{b}xd\left({F}_{i}^{j}\left(x\right)\right){\rm{d}}x}}-{\text{e}}^{{\displaystyle {\int }_{a}^{b}xd\left({F}_{k}^{j} \left(x\right)\right){\rm{d}}x}}}{{\text{e}}^{{\displaystyle {\int }_{a}^{b}xd\left({F}_{i}^{j} \left(x\right)\right){\rm{d}}x}}}\text{，}& {F}_{i}^{j}\left(x\right)\text{MD}{F}_{k}^{j}\left(x\right)\text{，}\\ 0\text{，}& {\rm{others}}\text{。}\end{array}\right.$ (12)

1）当且仅当 $F_i^j\left( x \right) \ne F_k^j\left( x \right)$ ，且 ${H_1}\left( x \right) = F_i^j\left( x \right) - F_k^j\left( x \right) \leqslant 0$ ，则称 $F_i^j\left( x \right)$ 一阶混合占优于 $F_k^j\left( x \right)$

2）当且仅当 $F_i^j\left( x \right) \ne F_k^j\left( x \right)$ ，且 ${H_2}\left( x \right) = \int_a^x {{H_1}\left( y \right)dy} \leqslant 0$ ，则称 $F_i^j\left( x \right)$ 二阶混合占优于 $F_k^j\left( x \right)$

3）当且仅当 $F_i^j\left( x \right) \ne F_k^j\left( x \right)$ ，且 ${H_3}\left( x \right) = \int_a^x {{H_2}\left( y \right)dy} \leqslant 0$ ，则称 $F_i^j\left( x \right)$ 三阶混合占优于 $F_k^j\left( x \right)$

3.3 方案择优

 ${\varPhi ^ + }\left( {{X_i}} \right) = \frac{1}{{m - 1}}\sum\limits_{k = 1}^m {{q_{ik}}}\text{，}$ (13)
 ${\varPhi ^ - }\left( {{X_i}} \right) = \frac{1}{{m - 1}}\sum\limits_{k = 1}^m {{q_{ki}}} \text{，}$ (14)
 $\varPhi \left( {{X_i}} \right) = {\varPhi ^ + }\left( {{X_i}} \right) - {\varPhi ^ - }\left( {{X_i}} \right) \text{。}$ (15)

4 实例应用 4.1 评价对象

4.2 指标属性值量化

4.2.1 绿色性、智能性指标

4.2.2 技术性、社会效益性指标

4.2.3 经济性指标

1）船价的确定

 $f\left( x \right) = \frac{1}{{0.01x\sqrt {2\text{π} } }}{e^{{\text{ - }}\frac{{{{\left( {\ln x - 0.2} \right)}^2}}}{{{\text{0}}{\text{.0002}}}}}} \text{。}$ (16)

2）燃油价格

 $f\left( x \right) = \frac{{2.85}}{{488.33}}{\left( {\frac{x}{{488.33}}} \right)^{2.85 - 1}}{e^{ - {{\left( {\frac{x}{{488.33}}} \right)}^{2.85}}}} \text{。}$ (17)

3）绿色智能技术成本增量

4）必要运费率的计算

 $RFR = \frac{{AAC}}{Q} = \frac{{\left( {P - L} \right) * \left( {\frac{A}{P},i,n} \right) + L * i + Y}}{Q} \text{。}$ (18)

5）计算结果

 图 3 经济性指标概率分布图 Fig. 3 Probability distribution of economic index

4.3 多属性决策

 ${\boldsymbol{C}} = \left[ {\begin{array}{*{20}{c}} 0 & {0.0166} & {0.0304} & {0.0314} & {0.0722} & {0.0707} \\ {0.0030} & 0 & {0.0331} & {0.0229} & {0.0746} & {0.0618} \\ {0.0369} & {0.0527} & 0 & {0.0114} & {0.0470} & {0.0478} \\ {0.0423} & {0.0465} & {0.0169} & 0 & {0.0626} & {0.0624}\\ {0.0811} & {0.0969} & {0.0485} & {0.0599} & 0 & {0.0252}\\ {0.0680} & {0.0730} & {0.0485} & {0.0485} & {0.0149} & 0 \end{array}} \right]。$ (19)

5 结　语

 [1] ZHANG Yanlong, WANG Lizheng. Comprehensive evaluation index system of green intelligent technology for Yangtze River ships[C]// International offshore(Ocean) and Polar Engineering Conference. 2020. [2] 张矢宇, 杨宇昊, 汪敏, 等. 长江绿色船舶评价指标体系研究[J]. 水运管理, 2021, 43(7): 24-27. DOI:10.13340/j.jsm.2021.07.008 [3] 喻露, 王志芳, 肖曙明, 等. 内河绿色船舶综合定量评价体系[J]. 中国航海, 2021, 44(1): 126-131. [4] 陈小卫, 王文双, 宋贵宝, 等. 基于模糊偏序关系的混合型多属性决策方法[J]. 系统工程与电子技术, 2012, 34(3): 529-533. [5] 张尧, 樊治平. 具有多种信息形式的随机多属性决策方法研究[J]. 运筹与管理, 2011, 20(4): 69-76. [6] 王芃, 文福拴, 王斐, 等. 基于混合多属性评价的错峰用电预案编制方法[J]. 电力系统自动化, 2016, 40(5): 54-61+70. [7] 中国船级社. 内河绿色船舶规范[S]. 2020. [8] 中国船级社. 绿色生态船舶规范[S]. 2020. [9] 中国船级社. 智能船舶规范[S]. 2020. [10] 张彦隆. 长江绿色智能船舶指标体系构建及应用研究[D]. 武汉: 武汉理工大学, 2021. [11] 张峰辉. 基于语言偏好信息的多属性群决策方法[D]. 东南大学, 2019. [12] 徐泽水. 一种不确定型OWA算子及其在群决策中的应用[J]. 东南大学学报(自然科学版), 2002(1): 147-150. [13] 苏绍娟. 沿海干散货船舶运输的不确定性分析方法研究及应用[D]. 武汉: 武汉理工大学, 2007. [14] HALOUANI N, CHABCHOUB H, MARTEL J M. PROMETHEE-MD-2T method for project selection[J]. European Journal of Operational Research, 2007, 195(3):