基于对立学习的改进CoDE算法

A modified CoDE algorithm with opposition learning

针对复合差分进化(CoDE)算法中所采用策略的局部搜索能力较弱及种群初始个体性能较差的缺点,提出了一种改进的CoDE算法.采用对立学习方式对种群进行初始化,改善初始解的性能;为加强算法的局部开发能力,提出了一个具有精英解的新变异策略以改进CoDE变异策略池.对30个CEC2014测试函数进行数值模拟并与9种算法进行了比较,结果表明该算法提高了计算的收敛速度和精度.

To alleviate the drawbacks of CoDE whose chosen strategies were weak exploitative and the initial population always had relatively poor performance,the opposition learning mechanism was first employed to improve the performance by initializing the population.A new mutation strategy with an elite solution was proposed to replace with the mutation strategy in CoDE,such that the exploitation of the algorithm was enhanced.A comparison of the proposed algorithm with nine famous algorithms was conducted on 30 CEC2014 test functions to evaluate its performance.The experimental results showed that the proposed algorithm was very competitive.