摘要
为了提高基本差分进化算法的寻优速度和寻优效能,提出了一种改进的自适应差分进化算法(ADE)。在基本差分进化算法中引入了自适应变异算子,根据每个个体与最优个体适应度值的相互关系,自动地调节变异算子值,使之在进化初期较大,随着个体逐渐接近最优值,算子值逐渐变小,确保个体向最优值快速、稳定地逼近。在每一代变异、交叉和竞争之后,又增加了与随机新种群的竞争操作,使算法易于跳出局部最优点,以提高全局搜索能力。采用4个经典的测试函数对算法进行验证,结果显示:该算法的收敛速度与收敛精度在一定程度上优于基本差分进化算法,同时也优于基于代数进行自适应变异的差分进化算法。
By using a new adaptive mutation operator, this paper proposes a modified adaptive differential evolution （ADE） algorithm to improve the optimum speed and performance of the differential evolution algorithm. The mutation operator is adjusted by the relationship between every individualrs fitness and the best one＇s fitness. The value of mutation operator is bigger at the beginning of the evolutionary process and will become smaller as the individual tending the optimal solution so as to quickly and stably approximate the best individual. After every basic mutation, crossover and competition, a new competition with a random swarm is added so as to effectively jump out of the local optimum and enhance the ability of global search. The simulation results for four classic functions show that both the convergence speed and accuracy of ADE are significantly superior to the differential evolution （DE） algorithm and the adaptive differential evolution algorithm that is based on the generation.
出处
《华东理工大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第4期600-605,共6页
Journal of East China University of Science and Technology
基金
国家杰出青年科学基金(60625302)
国家973项目(2009CB320603)
高等学校博士学科点专项科研基金新教师基金项目(200802511011)
长江学者和创新团队发展计划(IRT0721)
高等学校学科创新引智计划(B08021)
上海市重点学科建设项目(B504)
关键词
差分进化
交叉
变异
竞争
自适应
differential evolution
crossover
mutation
competition
adaptive
