摘要
针对带有边界变量的非线性优化问题,提出了一种改进的进化规划算法。该算法首先把每个个体看作为带有不同质量的粒子,根据目标函数值定义个体的质量。基于聚类思想,选取一定数量的点,然后利用选取的点分别求出于每个点相对应的重心,以每个点与其重心的连线方向为变异方向。最后把该方法应用到几个典型数值例子中,并与基本进化算法进行比较,数值结果表明算法是可行的、有效的。
In this paper, an improved evolutionary programming algorithm is proposed for nonlinear optimization problems with boundary variables. In this algorithm, each individual is taken as a particle, having different qualities. The qualities are defined according to their objective function values. Based on idea of clustering, some individuals are selected. The bar center of the selected individuals is calculated. The lines, through the individuals and their bar center, are the mutation's directions. In the end of the paper, the improved algorithm is applied to the some standard test functions and compared with the standard evolutionary programming algorithms. The numerical results demonstrate that the improved algorithm is efficient and practical.
出处
《辽宁工程技术大学学报(自然科学版)》
CAS
北大核心
2008年第1期155-157,共3页
Journal of Liaoning Technical University (Natural Science)
基金
辽宁省教育厅科学研究计划资助项目(2004C058)
关键词
进化规划
重心
随机搜索
全局最优解
非线性规划
evolutionary programming
bar center
random search
global optimization solution
nonlinear programming