An Efficient Optimization Method for HCC Antennas Using Quantum Genes Based on Sparse Artificial Neural Networks

An improved quantum genetic algorithm (IQGA) is proposed to optimize the antenna with a high computational cost (HCC) characteristic. First, the common problems in optimizing HCC antennas are analyzed. Then, a quantum gene is proposed to express antenna schemes. The bit of the quantum gene does not store 0 or 1 but the superposition state of 0 and 1. Therefore, the quantum gene can express more antenna schemes than the ordinary gene with the same gene size, which can significantly improve the algorithm’s searchability with a small population size. In quantum computing, an improved quantum rotation gate operator is designed. Since performing an EM simulation for HCC antennas is too time-consuming, a sparse artificial neural network (SANN) is proposed to increase the antenna scheme evaluation speed, thus reducing the number of EM simulations called by the algorithm. The real-world antenna experiment and the benchmark experiment show that IQGA proposed in this article can efficiently optimize the HCC antenna in a small population size.