适于电磁场逆问题鲁棒和全局优化的蚁群算法研究  

Robust and global optimizer based on ant colony algorithm for inverse problems in electromagnetics

作　　者：聂曼[1] 杨仕友[1] 

机构地区：[1]浙江大学电气工程学院,浙江杭州310027

出　　处：《电工电能新技术》2014年第3期44-47,70,共5页Advanced Technology of Electrical Engineering and Energy

摘　　要：为克服现有鲁棒优化算法计算效率低的问题,本文提出了一种适于电磁场逆问题鲁棒和全局优化的蚁群(ACO)快速优化算法。算法重新设计了算法结构,提出了有效计算鲁棒性能参数的方法。数值优化结果验证了算法的优越性和有效性。In real world design problems, the optimized solution is often very sensitive to small variations of the op- timized decision parameters. Therefore, it is important to apply robust optimal techniques to find the optimal solu- tion which has good performances in both objective function and robustness in the designed parameters against small perturbations. However, existing robust optimal techniques have difficulties in efficiently finding final solutions of inverse electromagnetic problems. In order to eliminate this problem, a fast ant colony optimization （ACO） algo- rithm based robust and global optimizer is proposed. In the proposed algorithm, the structures of the algorithm are redesigned, and methodologies to efficiently compute the robust performance parameters are proposed. The numeri- cal optimization results of a case study showcase the merits of the proposed algorithm, and validate the effectiveness of the proposed algorithm by comparing to the performance of different algorithms.

关 键 词：蚁群算法 电磁场逆问题 优化 鲁棒设计 

分 类 号：TM133[电气工程—电工理论与新技术]