一种基于拟态物理学优化的多目标优化算法

提出一种使用拟态物理学优化(APO)解决多目标优化问题的算法(MOAPO).根据多目标优化问题的特点,借鉴聚集函数法的思想,利用APO算法实现了对多目标优化问题中Pareto最优解集的搜索,并且在搜索过程中动态调整惯性权重与引力因子,以增强非劣解的多样性.实验结果表明了将APO应用于多目标优化问题的有效性.通过与基于微粒群优化(PSO)的多目标优化算法及NSGA-Ⅱ算法的比较,表明了MOAPO算法具有较好的分布性.

Multi-objective optimization algorithm based on artificial physics

A multi-objective optimization algorithm based on artificial physics optimization (MOAPO) is presented to
solve multi-objective optimization problems. According to the trait of multi-objective problems, by drawing lessons from
aggregating functions method, searching for Pareto optimal set of multi-objective optimization problems is implemented by
using APO algorithm. The inertia weight and gravitation coefficient are dynamic changing to explore the search space more
efficiently. The experimental simulations show that MOAPO is effective for multi-objective problems with a better diversity
compared with NSGA-II algorithm and multi-objective optimization algorithms based on praticle swarm optimization (PSO).