摘要
提出了罚函数与修复的混合策略,改进了罚函数,给出了种群早熟度评价指标,对遗传算法的交叉算子进行了改进,解决了单一使用罚函数方法求约束优化问题所遇到的困难,方便了遗传算法在约束优化问题中的应用,提高了遗传算法在机械及工程中应用的适应性。数值实验证明,该方法比传统的遗传算法处理约束优化问题效率高。
A mixed strategy of penalty function and repair was proposed and the penalty function was improved.The precocious degree evaluation index of population was given.The crossover operator of genetic algorithm was improved and the difficulties were solved which by only using penalty function to solve constrained optimization problem.This facilitated the genetic algorithm in using of constrained optimization problem,and improved the adaptability of genetic algorithm in the application of mechanical and engineering.Numerical experiments show that the method is efficient than traditional genetic algorithm in dealing with constrained optimization problems.
出处
《化工自动化及仪表》
CAS
北大核心
2010年第7期13-16,共4页
Control and Instruments in Chemical Industry
关键词
约束优化
罚函数
修复
遗传算法
constrained optimization
penalty function
repair
genetic algorithm
