摘要
针对约束优化问题的特点,给出一种改进差分进化算法.将差分进化算法做了适当修正,在初始化中加入迁移操作,依据违反约束度函数和原目标函数进行选择操作,当个体的违反约束度在容忍度以外时,通过违反约束度函数更新个体,当个体的违反约束度在容忍度以内时,通过原目标函数更新个体,实验研究结果表明该算法能有效求解约束优化问题.
For constrained optimization problem, an improved differential evolution algorithm is given. In the algorithm, differential evolution algorithm is appropriately modified. Migration operation is used in initialize. According to constraint violations function and original objective function selected operate. When the individual degree of violation constraints beyond the tolerance, the individual upda- ted by constraint violations function. When the individual degree of violation constraints within the tolerance, the individual updated by objective function. Numerical simulations show that this algo- rithm is effective for solving constrained optimization problems.
出处
《武汉理工大学学报（交通科学与工程版）》
2012年第5期1083-1087,共5页
Journal of Wuhan University of Technology（Transportation Science & Engineering）
基金
国家自然科学基金项目资助(批准号:60962006)
关键词
约束优化
差分进化算法
动态目标
迁移操作
constrained optimization
differential evolution algorithm
dynamic objective
migration operation
