Abstract
To eliminate the deficiency of traditional evolutionary algorithms in numerical implementations with respect to the three key operators such as selection, crossover and mutation operators, a new evolutionary algorithm based on probabilistic models, the quantum-inspired evolutionary algorithm (QEA), is studied for design optimizations of electromagnetic devices. In the proposed QEA, an adaptive update formulation is proposed for the rotation angle to balance the exploration and exploiting searches while the two level information sharing is simplified to one level one to facilitate the implementation of the proposed algorithm. Also, a new information sharing mechanism is introduced. The proposed algorithm is evaluated on an engineering inverse problem with promising results.