Abstract: To solve the problem of dual-frequency aperiodic array pattern synthesis, an optimization algorithm that combines genetic algorithm and differential evolution algorithm is proposed(GA-DE hybrid optimization algorithm). The optimization algorithm firstly uses the differential evolution algorithm to generate a population of the same size as the genetic algorithm in each iteration, mixes the population with the original population and retains individuals with better fitness, and then uses the genetic algorithm to iteratively optimize. At the same time, the selection, crossover and mutation operators of the genetic algorithm are improved, which improves the global search ability and convergence speed. Compared with other algorithms, the hybrid algorithm can be optimized to obtain better sidelobe level and gain. This simulation results show that the algorithm has good performance. Under the condition of ensuring the array gain, the hybrid optimization algorithm is used to optimize the Ku/ Ka dual-frequency mixed normalized subarray and Ku/ Ka dualfrequency mixed whole array with the goal of reducing the peak sidelobe level. After optimization, the maximum peak sidelobe levels of the Ku and Ka arrays in the scanning range are -13. 10 dB and -13. 13 dB respectively, and the minimum gains are 44. 81 dB and 44. 58 dB respectively, which meet the index requirements. The optimization results shows that the hybrid algorithm can effectively suppress the appearance of the aperiodic array grating lobes and reduce the peak sidelobe level of the array.