主观咨询约束下主客观一体化最优赋权方法

doi:10.12305/j.issn.1001-506X.2025.09.21

摘要/Abstract

摘要：

针对乘法合成法、加法合成法等赋权方法存在的主客观片面化、权重赋值关联性大、主观咨询约束缺乏等问题，提出一种主观咨询约束下主客观一体化赋权优化方法。分析主观约束类型与约束形态内涵，建立调节因子主观咨询约束模型，基于主成分分析（principal component analysis，PCA）法、逼近理想解排序法（technique for order preference by similarity to ideal solution，TOPSIS）、遗传算法（genetic algorithm，GA）等思想，设计调节因子种群适应度。相较于乘法合成法和加法合成法，所提方法计算的指标权重冗余度最大降低29.1%，准确度最大提高66.6%。所提方法赋权结果更好，同时具有更高的扩展性。

关键词: 主观咨询约束, 主客观一体化, 最优权重, 惩罚因子

Abstract:

Aiming at the problems of subjective and objective sidedness, high correlation between weight assignment, and lack of subjective consultation constraints in weighting methods such as multiplication synthesis and addition synthesis, a subjective and objective integration weighting optimization method with subjective consultation constraints is proposed. Analyze the subjective constraint type and constrained form connotation, establish a subjective consultation constraints model for regulatory factors, and design the population fitness of regulatory factors based on principal component analysis (PCA), technique for order preference by similarity to ideal solution (TOPSIS), genetic algorithm (GA), and other ideas. Compared with the multiplication synthesis method and the addition synthesis method, the proposed method reduces the redundancy of indicator weights by up to 29.1% and improves accuracy by up to 66.6%. The proposed method has better weighting results and higher scalability.

Key words: subjective consultation constraint, subjective and objective integration, optimal weight, penalty factor

中图分类号:

E 917

孙旭涛, 付勇, 王伟超. 主观咨询约束下主客观一体化最优赋权方法[J]. 系统工程与电子技术, 2025, 47(9): 3004-3010.

Xutao SUN, Yong FU, Weichao WANG. Optimal weighting method for subjective and objective integration with subjective consultation constraints[J]. Systems Engineering and Electronics, 2025, 47(9): 3004-3010.

图/表 7

表1

表2

表3

图1

图2

表4

表5

参考文献 31

1	JARROTT E, JURIS J, TURNER G. The intergenerational program evaluation tool: results of a delphi study[J]. Journal of Intergenerational Relationships, 2023, 23 (3): 280- 296.
2	LIN S C. Comparison of methods to derive relative weights.[J]. International Journal of Applied Mathematics, 2023, 53 (2): 507- 517.
3	XU Q H, TIAN B Y. Research on the method of power grid investment considering the difference and common points of binomial coefficient and variation coefficient[J]. Academic Journal of Business & Management, 2023, 5 (2): 167- 170.
4	ZHANG H W, WANG H Y, YAN Y S. Remote ship detection using relative multiscale weighted link entropy in marine environment[J]. Ocean Engineering, 2024, 295, 116976.
5	MARTINEZ S, ASTRID J. Selecting the distribution system using AHP and fuzzy AHP methods[J]. Mobile Networks & Application, 2024, 29, 235- 242.
6	ASLAN V. Determination of van basin groundwater potential by GIS based on AHP and Fuzzy-AHP methods.[J]. Journal of Agricultural Sciences, 2024, 30 (1): 47- 60.
7	VICTORIA M, MILOS M, EMILIO L. Selection of an international distribution center location: a comparison between stand-alone ANP and DEMATEL-ANP applications[J]. Research in Rransportation Business& Management, 2024, 56 (1): 101135.
8	NALBANT K, SAHIKA K, YAVUZ O. Evaluating the campus climate factors using an interval type-2 fuzzy ANP.[J]. Journal of Engineering & Natural Sciences, 2024, 42 (1): 89- 98.
9	GAO Y L, FENG Y Z, XIE Y W. Normalized robust PCA with adaptive reconstruction error minimization[J]. IEEE Trans. on Knowledge and Data Engineering, 2024, 36 (6): 2587- 2599.
10	LI X Y, LI C C, GUO F. Coefficient of variation method combined with XGboost ensemble model for wheat growth monitoring[J]. Frontiers in Plant Science, 2024, 14, 1267108.
11	HALICKA K, GOLA A. Gerontechnology ranking using the TOPSIS methods[J]. Engineering Management in Production and Services, 2024, 16 (1): 93- 103.
12	LI Y, CAI Q, WEI G W. Grey relational analysis method for multi-attribute group decision making based on cumulative prospect theory under single-value neutrosophic sets.[J]. Journal of Intelligent& Fuzzy Systems, 2024, 46 (1): 805- 819.
13	DEQUAN J, MURONG Y, QIN Z Y. A weighting method for feature dimension by semisupervised learning with entropy[J]. IEEE Trans. on Neural Networks and Learning Systems, 2023, 34 (3): 1218- 1227.
14	XIN S, XIANG G, TAO S. A novel subset assignment and matching method for DIC measurement of complex deformation in aircraft braking process[J]. Chinese Journal of Aeronautics, 2024, 37 (8): 373- 387.
15	SILGUERA J. Evaluation of mean square errors in simple moving average versus exponential smoothing method and assessment of time as predictor in forecasting myocardial infarction cases in the philippines[J]. International Journal of Multidisciplinary: Applied Business & Education Research, 2024, 5 (4): 1228- 1239.
16	HAO L, XIN G, LEI Z. Direct synthesis method of wideband tunable filtering phase shifter based on complex matched transversal multiple resonances[J]. International Journal of Electronics and Communications, 2024, 176 (1): 155151.
17	VIN L, SONG H Y, MINH T P. Synthesis of compliant parallel mechanisms using an improved beam-based method with the optimization of multiple resonant modes[J]. Machines, 2023, 11 (7): 731.
18	李刚. 主客观权重的组合方式及其合理性研究[J]. 复印报刊资料（管理科学）, 2018, 1 (3): 23- 33.
	LI G. Research on the combination of subjective and objective weights and their rationality[J]. Coping with Newspaper Materials （Management Science）, 2018, 1 (3): 23- 33.
19	TIAN X R, NING J, JIN C J. Analysis of spatial and temporal variation in sustainable water resources and their use based on improved combination weights[J]. Water, 2023, 15 (15): 2752.
20	LI W Y, DONG F G, JI Z S. Evaluation of provincial power supply reliability with high penetration of renewable energy based on combination weighting of game theory-TOPSIS method[J]. Sustainable Energy, Grids and Networks, 2023, 35 (1): 101092.
21	ZHANG Z, WANG Y Y, WANG Z X. A grey TOPSIS method based on weighted relational coefficient（Article）[J]. Journal of Grey System, 2014, 26 (2): 112- 123.
22	ZHANG J X, HUANG D Y, YOU Q J. Evaluation of emergency evacuation capacity of urban metro stations based on combined weights and TOPSIS-GRA method in intuitive fuzzy environment[J]. International Journal of Disaster Risk Reduction, 2023, 95 (1): 103864.
23	SONG X W, GUO S L, HAN L N. Research on hair removal algorithm of dermatoscopic images based on maximum variance fuzzy clustering and optimization Criminisi algorithm[J]. Biomedical Signal Processing and Control, 2022, 78 (1): 103967.
24	ZHOU H A. Model and algorithm based on sum of deviation squares for multi-criteria decision-making[J]. Material Science, Civil Engineering and Architecture Science, Mechanical Engineering and Manufacturing Technology, 2014, 651/653, 1603- 1606.
25	LO T P, GUO S J. Effective weighting model based on the maximum deviation with uncertain information[J]. Expert Systems with Applications, 2010, 37 (12): 8445- 8449.
26	马光, 朱文, 顾慧杰, 等. 基于加权最小绝对值状态估计的有源配电网拓扑辨识方法[J]. 中国电力, 2024, 57 (1): 167- 174.
	MA G, ZHU W, GU H J, et al. Topology identification method for active distribution network based on weighted minimum absolute value state estimation[J]. Electric Power, 2024, 57 (1): 167- 174.
27	WANG C L, GAO A, XUAN J P. Optimal demodulation band extraction method for bearing faults diagnosis based on weighted geometric cyclic relative entropy[J]. Machines, 2023, 11(1): 39.
28	ZENG W, YI G D, ZHANG S Y. Multi-objective optimization method of injection molding process parameters based on hierarchical sampling and comprehensive entropy weights[J]. International Journal of Advanced Manufacturing Technology, 2024, 133 (3): 1481- 1499.
29	ZHANG D Z, ZHANG T L, JIA S C. Tuning synaptic connections instead of weights by genetic algorithm in spiking policy network[J]. Machine Intelligence Research, 2024, 1 (5): 906- 918.
30	谢婼青. 专家咨询约束下综合评价指标体系构建的研究及应用[D]. 上海: 上海社会科学院, 2019.
	XIE R Q. Research and application of comprehensive evaluation index system construction under expert consultation constraints[D]. Shanghai: Shanghai Academy of Social Sciences, 2019.
31	谈宜东, 徐欣, 张书练. 激光干涉精密测量与应用[J]. 中国激光, 2021, 48 (1): 228- 249.
	TAN Y D, XU X, ZHANG S L. Laser interference precision measurement and application[J]. China Laser, 2021, 48 (1): 228- 249.

指标	指标权重赋值
指标	$ {x_{^1}} $	$ {x_2} $	…	$ {x_L} $
$ {v_{^1}} $	$ {x_{11}} $	$ {x_{12}} $	…	$ {x_{1L}} $
$ {v_2} $	$ {x_{21}} $	$ {x_{22}} $	…	$ {x_{2L}} $
$\vdots $	$\vdots $	$\vdots $		$\vdots $
$ {v_p} $	$ {x_{p1}} $	$ {x_{p2}} $	…	$ {x_{pL}} $

分类条件	分类结果
$ {L_{^1}} > 0 \wedge {L_{^2}} = 0 \wedge {L_{^3}} = 0 $	客观一体化最优权重
$ {L_{^1}} = 0 \wedge {L_{^2}} > 0 \wedge {L_{^3}} = 0 $	主观一体化最优权重
$ {L_{^1}} = 0 \wedge {L_{^2}} = 0 \wedge {L_{^3}} > 0 $	组合赋权一体化最优权重
$ {L_{^1}} > 0 \wedge {L_{^2}} > 0 \wedge {L_{^3}} = 0 $	主客观混合一体化最优权重
$ {L_{^1}} > 0 \wedge {L_{^2}} > 0 \wedge {L_{^3}} > 0 $	主客观与组合赋权一体化最优权重

指标	指标权重赋值						指标约束类型
指标	$ {x_{^1}} $	$ {x_{^2}} $	$\cdots $	$ {x_L} $	$ \alpha {{{\alpha }}^ +_i }/\left\\| {{{\boldsymbol{\alpha }}^ + }} \right\\| $	$ \beta {{{\beta }}^ -_i }/\left\\| {{{\boldsymbol{\beta }}^ - }} \right\\| $	指标约束类型
$ {v_{^1}} $	$ {x_{11}} $	$ {x_{12}} $	$\cdots $	$ {x_{1L}} $	$ \alpha \alpha _1^ + /\left\\| {{\boldsymbol \alpha ^ + }} \right\\| $	$ \beta \beta _1^ - /\left\\| {{\boldsymbol \beta ^ - }} \right\\| $	$ U({v_1}) $
$ {v_2} $	$ {x_{21}} $	$ {x_{21}} $	$\cdots $	$ {x_{2L}} $	$ \alpha \alpha _2^ + /\left\\| {{\boldsymbol \alpha ^ + }} \right\\| $	$\beta \beta _2^ - /\left\\| {{\boldsymbol \beta ^ - }} \right\\| $	$ U({v_2}) $
$\vdots $	$\vdots $	$\vdots $		$\vdots $	$\vdots $	$\vdots $	$\vdots $
$ {v_p} $	$ {x_{p1}} $	$ {x_{p2}} $	$\cdots $	$ {x_{pL}} $	$ \alpha \alpha _p^ + /\left\\| {{\boldsymbol \alpha ^ + }} \right\\| $	$ \beta \beta _p^ - /\left\\| {{\boldsymbol \beta ^ - }} \right\\| $	$ U({v_p}) $

指标	$ {X_{^1}} $	$ {X_2} $	$ {X_3} $	$ {X_4} $	$ {X_5} $	$ {X_6} $	$ {X_7} $	$ {X_8} $	指标约束
电磁信号测试能力A₁	0.206	0.183	0.141	0.215	0.146	0.180	0.215	0.141	$ {W_{{A_1}}} \geqslant 0.18 $
网络安全测试能力A₂	0.152	0.152	0.198	0.120	0.175	0.103	0.198	0.103	$ {W_{{A_2}}} \geqslant 0.12 $
测试发射能力A₃	0.159	0.127	0.226	0.144	0.198	0.180	0.226	0.127	$ {W_{{A_3}}} < 0.25 $
测量控制能力A₄	0.129	0.178	0.113	0.167	0.146	0.180	0.180	0.113	$ {W_{{A_4}}} \geqslant 0.10 $
目标特性测量能力A₅	0.180	0.168	0.164	0.172	0.201	0.208	0.208	0.164	$ {W_{{A_5}}} < 0.20 $
毁伤效应测量能力A₆	0.175	0.193	0.158	0.182	0.134	0.149	0.193	0.134	$ 0.10 < {W_{{A_6}}} < 0.20 $

载荷系数				线性组合系数				综合评分系数
0.537	0.211	−0.707	0.089	2.294	1.573	−8.187	1.183	0.271
0.113	−0.448	−0.113	0.481	0.481	−3.341	−1.314	6.416	−0.196
0.472	−0.137	−0.033	−0.111	2.015	−1.020	−0.386	−1.486	0.323
0.393	−0.481	0.299	−0.580	1.677	−3.587	3.463	−7.734	0.320
−0.064	−0.624	−0.046	0.406	−0.271	−4.652	−0.534	5.415	−0.383
0.504	0.330	0.620	0.466	2.152	2.459	7.184	6.216	0.529
0.241	−0.044	−0.105	−0.152	1.030	−0.331	−1.214	−2.030	0.057
0.079	−0.059	0.008	0.077	0.336	−0.439	0.090	1.022	0.078