Abstract: Owing to the inconvenient installation and coating of traditional infrared absorbers, a double-layer quantum dot structure is proposed. An array was constructed on a gold substrate using this unit structure, and a near-infrared absorber was reverse-designed and optimized using an intelligent optimization algorithm. The inner and outer layers of the double-layer quantum dot were composed of gold and silicon dioxide, respectively. To save time and improve efficiency, a particle swarm optimization algorithm based on the double improvement of the chaotic mapping and Levy flight strategies was considered. The selected structural size was simulated, and the highest achievable absorption rate of the quantum dot array and its corresponding structural size were repeatedly searched. Finally, the optimal array was verified through simulation experiments. The results revealed an absorption peak with an absorption rate of 99.96% at 1159 nm, which achieved narrow-band perfect absorption in the near-infrared band. Moreover, the absorption effect was better than that of absorber structures obtained using other commonly used optimization algorithms. Perfect absorption of other frequency points may be achieved by regulating the structural parameters. This method provides a solution for the automated design of new optical devices.