Tabu search based robust design optimization for inverse problems in electromagnetics

CAO Ke-jian， YANG Shi-you， WU Lie
（College of Electrical Engineering， Zhejiang University， Hangzhou 310027， China）

Abstract： In order to present a comprehensive analysis of the performance of the optimal，the robust performance of optimal solutions was examined under the circumstance of uncertainties，which are always incurred in practical inverse problems in electromagnetics. Some robust design optimization methods were discussed and analyzed，especially the estimation of the key parameter in evaluating the robust performance. With the goal of finding the robust optimal solution，the improved tabu search algorithm was applied to a typical numerical problem and an engineering design problem. By computing and comparing the deviation of object function values in condition of a perturbation to the optimal solution，the results show that the robust performance parameter of the robust optimal solution is much stronger than that of the global optimal one.
Key words： robust optimized design；expected fitness function；tabu search algorithm；inverse problems