红外吸收常用于军事对抗等领域，具有重要研究意义。针对传统红外吸收器存在不便安装和涂敷的缺点，提出了一种双壳层量子点结构，其制备较为简单，涂敷方法多元。本文根据吸收器的技术指标，依托双壳层量子点结构，选取合适的算法对结构参数进行优化，实现了窄带吸收器和宽带吸收器的自动化设计。研究内容如下：

（1）提出了一种量子点阵列结构来设计窄带吸收器，并选用改进后的粒子群算法进行结构优化，实现了窄带完美吸收器的自动化设计。首先，为了克服粒子群算法容易陷入局部最优的问题，引入改进后的莱维飞行策略和混沌映射策略对其进行改进，得到改进后的粒子群算法(LFBC-PSO)。其次，设置量子点半径、介电层厚度、量子点间距这些参数的取值范围，根据算法迭代产生的结构参数在金衬底上排布二氧化硅包裹金的双壳层量子点材料，得到对应的结构。最后，根据LFBC-PSO算法反复迭代过程，搜索量子点阵列的最佳吸收率，并得到对应的结构参数。实验结果表明，该吸收器在900nm-1200nm波段的1159nm处有吸收率为99.96%的完美吸收峰，将波段范围改为1200nm-1500nm后，在1278nm处仍能找到吸收率为99.64%的完美吸收峰。这表明在LFBC-PSO算法对结构的的优化下，可以实现特定波长的光吸收，完成了窄带完美吸收器的逆向设计。同时，在原有量子点阵列的竖直方向上堆叠其他双层量子点阵列，对宽带吸收器的设计进行了初步探索，并实现了多频点完美吸收器的设计，在1064nm、1157nm、1200nm处均出现了吸收率超过95%的吸收峰。

（2）提出了一种复合量子点阵列结构来设计宽带吸收器，并选用改进后的麻雀算法进行结构优化，实现了宽带完美吸收器的自动化设计。其中，复合量子点阵列分为上下两层，上层的量子点阵列以二氧化硅包裹钛的双壳层量子点周期排布，下层的量子点阵列以二氧化硅包裹硅的双壳层量子点周期排布。首先，针对宽带吸收器设计这一多参数优化问题，利用位置更新策略、自适应t分布变异策略和混沌映射策略对麻雀算法进行改进，得到改进后的麻雀算法(NGTSSA)。其次，设置量子点半径、壳层厚度、量子点间距以及阵列间间距这些参数的取值范围，根据算法迭代产生的结构参数在FDTD中构建吸收器，得到对应的结构。并且通过改进后的麻雀算法对所设置的参数进行优化，挑选出宽带吸收效果好（平均吸收率高且吸收带宽宽）的结构参数。实验结果表明，该双层量子点阵列结构在1200-2000nm处实现了超宽带完美吸收，平均吸收率为90.11%，吸收效果远优于传统贵金属(Au、Ag)和耐高温金属(W、Cr)所组成的吸收器结构，完成了宽带完美吸收器的逆向设计。

Infrared absorption is often used in military confrontation and other fields, and has important research significance. In view of the shortcomings of traditional infrared absorbers that are inconvenient to install and coat, a double-shell quantum dot structure is proposed, which is relatively simple to prepare and has multiple coating methods. Based on the technical indicators of the absorber and relying on the double-shell quantum dot structure, this paper selects a suitable algorithm to optimize the structural parameters, and realizes the automated design of narrowband absorbers and broadband absorbers. The research contents are as follows:

(1)A quantum dot array structure is proposed to design a narrowband absorber, and an improved particle swarm algorithm is used for structural optimization to achieve the automated design of a narrowband perfect absorber. First, in order to overcome the problem that the particle swarm algorithm is prone to fall into the local optimum, the improved Levy flight strategy and chaotic mapping strategy are introduced to improve it, and the improved particle swarm algorithm (Lévy Flight and Chaotic Mapping-Based Particle Swarm Optimization，LFBC-PSO) is obtained. Secondly, the value ranges of the parameters such as quantum dot radius, dielectric layer thickness, and quantum dot spacing are set, and the double-shell quantum dot material wrapped in silicon dioxide is arranged on the gold substrate according to the structural parameters generated by the algorithm iteration to obtain the corresponding structure. Finally, according to the repeated iteration process of the LFBC-PSO algorithm, the optimal absorption rate of the quantum dot array is searched, and the corresponding structural parameters are obtained. The experimental results show that the absorber has a perfect absorption peak with an absorption rate of 99.96% at 1159nm in the 900nm-1200nm band. After changing the band range to 1200nm-1500nm, the perfect absorption peak with an absorption rate of 99.64% can still be found at 1278nm. This shows that under the optimization of the structure by the LFBC-PSO algorithm, light absorption of a specific wavelength can be achieved, completing the reverse design of a narrowband perfect absorber. At the same time, other double-layer quantum dot arrays were stacked in the vertical direction of the original quantum dot array, and the design of a broadband absorber was initially explored, and the design of a multi-frequency perfect absorber was achieved, with absorption peaks with an absorption rate of more than 95% appearing at 1064nm, 1157nm, and 1200nm.

(2)A composite quantum dot array structure is proposed to design a broadband absorber, and the improved sparrow algorithm is used for structural optimization to achieve the automated design of a broadband perfect absorber. The composite quantum dot array is divided into two layers, the upper quantum dot array is periodically arranged with double-shell quantum dots of titanium wrapped in silicon dioxide, and the lower quantum dot array is periodically arranged with double-shell quantum dots of silicon wrapped in silicon dioxide. Firstly, for the multi-parameter optimization problem of broadband absorber design, the sparrow algorithm is improved by using the position update strategy, adaptive t-distribution mutation strategy and chaotic mapping strategy to obtain the improved sparrow algorithm (NGTSSA). Secondly, the value ranges of the parameters such as quantum dot radius, shell thickness thickness, quantum dot spacing and array spacing are set, and the absorber is constructed in FDTD according to the structural parameters generated by the algorithm iteration to obtain the corresponding structure. And the set parameters are optimized by the improved sparrow algorithm to select the structural parameters with good broadband absorption effect (high average absorption rate and wide absorption bandwidth). Experimental results show that the double-layer quantum dot array structure achieves ultra-wideband perfect absorption at 1200-2000nm, with an average absorption rate of 90.11%. The absorption effect is far better than the absorber structure composed of traditional precious metals(Au,Ag)and high-temperature resistant metals(W,Cr), completing the reverse design of the broadband perfect absorber.