【论文解析】EJOR 2011 A clustering procedure for reducing the number of representative solutions in the Pareto Front of multiobjective optimization problems

发布时间 2023-07-09 10:17:25作者: 震灵

论文名称:A clustering procedure for reducing the number of representative solutions in the Pareto Front of multiobjective optimization problems

动机

假设一个三目标优化问题

\[\begin{aligned} & \text { Availability: } \max _\theta J_1(\theta)=\max _{\theta_p, p=1, \ldots, 5}\left[\prod_{p=1}^u A_p\left(\theta_p\right)\right] \\ & \text { Cost: } \quad \min _\theta J_2(\theta)=\underset{\theta_{p h}, p=1, \ldots, 5}{ }\left[\sum_{p=1}^u \sum_{h=1}^{m_p} c_{p h} \theta_{p h}\right] \\ & h=1, \ldots, m_p \\ & \end{aligned}\\ \text { Weight: } \quad \min _\theta J_3(\theta)=\min _{\substack{\theta_{p h}, p=1, \ldots, 5 \\ h=1, \ldots, m_p}}\left[\sum_{p=1}^u \sum_{h=1}^{m_p} w_{p h} \theta_{p h}\right] \]

Pareto Front上有很多解,毫无疑问,对于多目标决策来说,需要找到一个办法,找出其中的代表性解。
image

算法

Clustering是一个不错的思路。

image