哈尔滨工程大学学报Issue(9)：1179-1187,9.DOI:10.3969/j.issn.1006-7043.201303061
基于特征解选取的高维多目标可视化研究
The visualization of high dimension multi-objective based on the characteristic solution selection
摘要
Abstract
Current visualization techniques failed to effectively demonstrate the Pareto optimal solutions of high di-mensional multi-objective optimization problems. In view of this problem, a new graphical visualization technology of n-dimensional Pareto front based on the selection of characteristic solutions was proposed. First, solutions with obvious characteristics in the optimal solutions were collected in allusion to the feature of multi-objective optimiza-tion problems. After that, two objective information sharing mechanisms were proposed aiming at different decision needs and the characteristic solutions were sorted effectively after sharing objective information. Finally, a sub-graphic form was drawn. In this method, redundancy solutions of similar performance in Pareto optimal solutions were eliminated;objective information of characteristic solutions was sorted, the trend of performance and the deci-sion information of decision-making were visualized effectively. The multi-dimensional objective visualization model was designed based on the idea. This method achieves the purpose that the decision-maker can conveniently analyze the Pareto optimal solutions and make decisions.关键词
高维多目标/可视化技术/特征解选取/n维图表可视化/Pareto前沿决策Key words
high dimensional multi-objective/visualization technology/characteristic solution selection/n-dimen-sional diagrams visualization technology/P areto front decision分类
信息技术与安全科学引用本文复制引用
毕晓君,李博,王珏..基于特征解选取的高维多目标可视化研究[J].哈尔滨工程大学学报,2013,(9):1179-1187,9.基金项目
国家自然科学基金资助项目（61175126）；中央高校基本科研业务费专项基金资助项目（ HEUCFZ1209）；教育部博士点基金资助项目（20112304110009）． （）
