自适应并行遗传算法实现有源滤波器的设计  

Adaptive Parallel Genetic Programming Applied to Active Filter Design

作　　者：张志美[1] 程立英[1] 王洪刚[2] 孙博[1] 乔双[3] 

机构地区：[1]沈阳师范大学物理科学与技术学院,辽宁沈阳110034 [2]鲁东大学物理与电子工程学院,山东烟台264025 [3]东北师范大学物理学院,吉林长春130024

出　　处：《沈阳师范大学学报（自然科学版）》2008年第4期425-428,共4页Journal of Shenyang Normal University:Natural Science Edition

基　　金：国家高技术发展计划项目(2002AA632080)

摘　　要：提出一种改进的自适应并行遗传算法,通过设计自适应算子自动调节交叉和变异概率,采用基于岛屿交换模型实现多种群间信息交换,提高了种群的多样性和健壮性.将该方法用于滤波器的参数优化设计,给出4阶切比雪夫低通滤波器的设计结果,比较各种算法的性能,证明该算法收敛速度快,精度高,有效地克服了早熟现象.This paper introduces an improved adaptive parallel genetic algorithm and designs the adaptive operators to improve the probabilities of crossover and mutation automatically. The algorithm realizes information exchanges between multi populations based on the island model so to improve the diversity and robustness of populations. In the paper we applied this method to realizing a 4 orders Chebyshev low pass filter and compared other algorithms with this method. The simulation experiments show the final design results are good and can satisfy the requirements of precision and performance. The comparison experiments show this method has a higher convergence speed and can solve the problem of prematurity effectively.

关 键 词：自适应并行遗传算法 岛屿交换模型 有源滤波器 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]