基于多目标优化和多属性决策的数据采集装置总体优化设计  

Data acquisition unit optimal design based on multi-objective optimization and multi-attribute decision making

作　　者：杨路春[1] 胡迟 高阳 黄利华[1] 李学斌[1] YANG Lu-chun;HU Chi;GAO Yang;HUANG Li-hua;LI Xue-bin(Wuhan Second Ship Design and Research Institute,Wuhan 430064,China)

机构地区：[1]武汉第二船舶设计研究所,湖北武汉430064

出　　处：《舰船科学技术》2021年第1期164-169,共6页Ship Science and Technology

基　　金：装备发展部装备预研项目。

摘　　要：舰船数据采集装置的设计需兼顾装置总体质量、装置散热性能、装置结构变形及装置结构强度等多种要求,是一个典型的多目标优化设计问题。本文采用基于精英策略实数编码、带跳变算子的非支配排序多目标遗传算法(RNSGA-II-SBJG算法),开展考虑上述要求的数据采集装置总体优化设计;考虑数据采集装置工作过程中热-结构耦合效应,建立多场耦合分析模型,基于响应面方法构建代理模型,引入灰色关联度分析方法计算Pareto解集与理想解的关联度,对优化解集进行评价排序。应用上述方法对某数据采集装置进行优化设计,数值结果表明,径向基函数拟合精度高、Pareto解集分散良好,基于灰色关联度分析的优化解评价方法能够有效、合理地选取优化设计方案。The design of ship data acquisition unit is a typical multi-objective optimization problem since a proper trade-off is necessary among different aspects,such as the device total mass,cooling performance,device deformation and device strength.In this work,the design problem is solved by means of the RNSGA-II,a variant of the Real-coded Nondominated Sorting Genetic Algorithm(RNSGA).Considering the thermal-structure coupling effect during the working process of the unit,a multi-field coupling analysis model is established,and an agent approximate model is constructed based on radial basis functions(RBF)method.To determine the optimal design from the Pareto solution set,the grey relational analysis approach is employed.The Pareto solution set is ranked according to the degree of correlation between among alternative,the ideal and negative ideal solution.The proposed method is applied to optimize a data acquisition unit.The numerical results indicate that the RBF has good precision,the Pareto solutions are well dispersed,and the optimal solutions can be effectively identified by the grey relational analysis.

关 键 词：数据采集装置优化设计 多场耦合分析 响应面 RNSGA-II-SBJG 灰色关联度分析 

分 类 号：TN802[电子电信—信息与通信工程]