摘要
为了克服惩罚函数法存在的罚参数难以选择和控制的主要缺陷,利用个体违反约束条件的程度函数,定义了约束强度指标,并设计了一种新的具有较强全局搜索能力的多父体杂交算子,从而提出一种基于约束强度的有效的演化算法。通过数值验证比较其性能优于现有的一些约束单目标优化演化算法。
In order to overcome the major shortcomings of penalty function method that is difficult in choosing the penalty parameters and control, the degree of constraint of individual violation is used completly, the binding strength indicators are defined and a new strong global search ability of the father of many body crossover operator is designed, which presents a constraint based on the strength of effective evolutionary algorithm, through the validation to compare their performance is better than some of the existing constrained single-objective optimization evolutionary algorithms.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第7期1719-1721,共3页
Computer Engineering and Design
关键词
演化算法
约束单目标优化
约束强度
多父体杂交
evolutionary algorithms
constrained single-objective optimization
constraints strength
multi-parent crossover
