基于区间可信度下界的多目标优化算法研究及应用

发布时间：2018-01-31 19:04

本文关键词： 区间可信度占优关系多目标优化算法　出处：《计算机科学》2017年S2期 　论文类型：期刊论文

【摘要】：旨在填补多目标优化算法研究的不足,以制造业中多因素耦合作用下的多目标优化问题为研究对象,首先提出区间可信度和占优关系等概念;其次基于区间可信度和占优关系建立基于区间可信度下界的多目标优化算法;最后通过多目标数值优化对所建立的优化算法进行探究。结果显示,在γ的取值相同时,H测度与进化代数呈现正相关,可以说明随着个体进化代数的增大,所提出的基于占优可信度下界的算法得到的γ-Pareto前沿越能反映真实的Pareto前沿;通过文中建立的算法与IP-MOEA和SPGA的比较可以看出,文中所建立的基于区间可信度下界的多目标优化算法与实际情况的吻合度更高,说明所建立的算法可以填补多目标优化算法的不足。
[Abstract]:In order to fill in the deficiency of the multi-objective optimization algorithm, the multi-objective optimization problem under the multi-factor coupling in manufacturing industry is taken as the research object. Firstly, the concepts of interval reliability and dominant relation are put forward. Secondly, a multi-objective optimization algorithm based on the lower bound of interval confidence is established based on interval confidence and dominant relation. The result shows that the H measure is positively correlated with evolutionary algebra when the value of 纬 is the same, which shows that with the increase of individual evolutionary algebra, the optimization algorithm is studied by multi-objective numerical optimization. The 纬 -Pareto frontier based on the dominating confidence lower bound can reflect the real Pareto frontier more effectively. By comparing the algorithm with IP-MOEA and SPGA, it can be seen that the multi-objective optimization algorithm based on the lower bound of interval confidence is more consistent with the actual situation. It shows that the proposed algorithm can fill the deficiency of the multi-objective optimization algorithm.
【作者单位】：营口理工学院电气工程系;
【基金】：国家自然科学基金(51185319)资助
【分类号】：TG506;TP18
【正文快照】： 1引言随着计算机技术的飞速发展,机械制造业、汽车、航空航天以及其他各个行业都将依靠计算机技术对其进行控制。以机械制造业中难加工材料的高速切削加工为例进行说明,评价此种材料的切削加工性将以加工表面质量、加工效率、刀具磨损、切削力的大小以及切削温度等为目标。并

【相似文献】