国际中文开源期刊平台

基于偏好的区间多目标进化算法的比较与应用

Comparison and application of interval multi- objective evolutionary algorithm based on preference

应用数学资讯 / 2020,2(2)：35-44 / 2020-05-22 1971 3386

查看全文 下载文档

• 作者： 杜宁清     
• 单位：

  西安电子科技大学，西安

• 关键词： 进化算法；多目标优化；区间；偏好；机器人；路径规划
• evolutionary algorithm; multi-objective optimization; interval; preference; robot; path planning
• 摘要： 区间参数多目标优化问题是普遍存在且非常重要的。目前直接求解该 类问题的进化优化方法非常少，且已有方法的目的是找到收敛性好且分布均匀 的 Pareto 最优解集。为得到符合决策者偏好的最满意解，本文综述 3 种基于偏 好的区间多目标进化算法，并将其应用于特定环境下机器人路径规划问题，比 较 3 种算法的性能。研究结果可丰富特定环境下机器人路径规划的求解方法， 提高机器人路径优化效果。
• Interval multi-objective optimization problems are very popular and important. There exist few evolutionary optimization methods for directly solving them, and these existing methods aim at finding a set of well-converged and evenly- distributed Paretooptimal solutions. Three preference-based interval multi-objective evolutionary algorithms are surveyed to obtain the most preferred solution fitted the decision maker’s preferences. Additionally, the above algorithms are applied in robot path planning problems under a special environment, and are compared about their performance. The research enriches the methods of solving robot path planning problems under a special environment, and improves the optimization performance of the problems.
• DOI： https://doi.org/10.35534/ami.0202006c
• 引用： 杜宁清．基于偏好的区间多目标进化算法的比较与应用［J］．应用数学资讯，2020，2（2）：35-44．