研究进化算法求解动态偏好多目标优化问题，可以更好地满足决策者的动态需求。本项目针对动态偏好进化机制缺失问题，对动态偏好多目标优化进行系统的研究。针对现有MOP基准测试问题集不能同时兼具动态以及偏好两个方面性能的问题，提出具有代表性的引入动态变化的偏好信息多目标基准测试问题集；针对目前动态偏好算法中存在的无法快速收敛的问题，通过动态环境勘探技术和优化技术，构建动态响应策略，通过偏好角度设计、分解策略等方法跟踪决策者的偏好；针对进化DPMO领域还缺乏算法测试和评估体系这一急需解决的问题，进行评估动态偏好信息多目标进化效果的评价方法和实际应用效果研究，实现对优化结果的公平有效评价和算法的有效检测。通过以上关键问题的研究，建立动态偏好多目标优化的基准测试集、性能评价体系和基本框架，为动态偏好多目标进化优化算法的设计和应用提供新方法，为本领域的进一步研究和现实应用提供系统化工具和理论基础。

The research of evolutionary algorithms on solving dynamic preference-based multi-objective optimization (DPMO) problems can better satisfy decision makers’ all kinds of dynamic requirements. This project aims to systematically study dynamic preference-based multi-objective optimization to address the issue of shortage of dynamic preference-based evolutionary mechanism. To deal with the problems that most existing multi-objective benchmark problems cannot simultaneously possess the character of dynamics and preference, we will propose representative multi-objective optimization benchmark problems introducing varying preference information. As most present dynamic and preference-based optimization algorithms do not show fast convergence speed, we design dynamic response strategies by dynamic environmental exploring and optimizing techniques, and track decision makers’ preference information through the design of preference angle and decomposition strategies. Considering that there is a lack of system on algorithms testing and evaluating in the field of DPMO and its practical application effects, we study the evaluation methods to evaluate the evolutionary effectiveness of DPMO, achieving the fairly effective evaluation on the results of optimization and the high-efficiency test on optimization algorithms. By researching the above key issues, we will establish the benchmark problems of DPMO, performance assessment system and the general framework, providing novel approaches for the design and application of dynamic preference-based multi-objective evolutionary optimization algorithms, in order to provide systematical tools and theoretical basis for the further research and application in the domain.