摘要本课题采用帕累托机制的多目标优化算法,重点对 NSGA-II 展开研究,深入了解 NSGA-II 在遗传演化的选择过程中的算法原理,通过编程实现其算法流程,测试其优化 效果和性能。实现该算法的前提条件是定义并计算帕累托解的拥挤度及拥挤度比较算子。 本课题拟通过 Matlab 编程完成上述研究内容,并对实验结果实现可视化。通过实验结果 分析,NSGA-II 在运行时间、种群多样性等问题的处理上有较好的性能。以后对 NSGA-II 优化的研究方向包括:进一步提高运行效率,并使种群收敛更加均匀。78940
毕业论文关键词 帕累托最优 多目标优化 NSGA-II
Title The research on the implementation of NSGA-II Algorithm
Abstract The research task of this project is on the muti-objective optimization algorithm Based on Pareto mechanism, with emphasis on the study of NSGA-II algorithm。 Focusing on a deep understanding of the algorithmic principles on the selection of genetic evolution, this project verifies the optimality and performance of NSGA-II by implementing its algorithmic flow。 The prerequisite of the algorithm implementation is to definite and calculate crowding distance and congestion comparison operator。 The above-mentioned research task was accomplished by Matlab programming, and the experimental results were visualized and presented by a Java interface。 According to the analysis of results, NSGA-II achieves better results in computation time and species persity compared with NSGA。 The future research directions includes further improving the computational efficiency and achieving more evenly distributed species。
Keywords Pareto-Optimality Muti-objective Optimization NSGA-II
目 次
1 引言 1
1。1 应用背景 1
1。3 本文主要工作 2
1。4 本文结构 2
1。5 本章小结 2
2 相关预备知识 3
2。1 Pareto 最优的概念 3
2。2 传统的多目标优化方法 3
2。3进化算法(Evolutionary algorithms) 4
2。4多目标遗传算法(Genetic Algorithm) 5
2。5 本章小结 6
3 NSGA-II 算法的流程设计与实现 7
3。1 设计原理和设计目标 7
3。2 总体设计框架 7
3。3 NSGA-II 算法的详细流程 9
3。4 程序框架及主要数据结构 15
3。5 本章小结 19