一种改进的自适应差分进化算法

A Modified Adaptive Differential Evolution Algorithm

差分进化算法(DE)是一种基于群智能的全局搜索优化算法,具有可控参数少,实现简单,全局搜索能力强等优点,近年来引起来广泛的关注,并在多领域得到了成功应用。但该算法局部搜索能力不足,容易陷入局部最优解,易出现"早熟"现象。针对此问题提出一种改进的自适应差分进化算法(MADE),该算法基于进化个体的适应度值的优劣,对不同个体采取不同的变异策略,区别对待不同的优劣程度的解,从而提高求解效率。通过典型的优化函数对算法进行测试验证,结果表明所提算法相对于其他算法具有迭代次数少,求解精度高等优点。

Differential evolution algorithm（DE） is a kind of global search optimization algorithm based on swarm intelligence, which has mang advantages such as, less parameters, strong global search ability, and easy to implement. In recent years, it has been successfully applied in many fields. But it also faces the problem of low local search ability, easy to fall into local optimal solution and difficult to solve multimodal optimization problem effectively. Aimed at solving this problem, an adaptive differential evolution algorithm was proposed（MADE）. In this algorithm, different crossover and mutation strategy was applied based on the fitness value of individual. Some typical multi-objective optimization functions are tested to verify the algorithm, simulation results showed that MADE has obvious advantages than other algorithms, such as less iteration time.