基于插值方法的EMO多样性保持策略

Multi-object Optimization of Diversity Preservation Strategy Based on Interpolation-extrapolation Strategy

针对演化多目标优化(evolutionary multi-objective optimization,EMO)算法搜索到的近似Pareto前沿出现间断或不完整现象的问题,提出了基于插值方法的演化多目标多样性保持策略,利用标准测试函数对NSGAII(支配关系排序的遗传算法)和提出的算法进行多样性和收敛性指标测试,数值试验结果表明,所提出的算法性能得到明显改进,优于NSGAII算法。

There exist the following two situations:approximate Pareto front disconnects or the non-dominated solutions concentrate in some regions.The interpolation-extrapolation strategy was used to improve the search capabilities and to find more non-dominated solutions in those regions.The standard test function was used to measure the convergence and diversity of the presented algorithm and NSGAII algorithm.The numerical experiments show that the performance of the proposed algorithm has been improved significantly better than NSGAII algorithm.