The Goos–Hänchen (GH) shift is one of the important aspects to evaluate the performance of multilayer surface plasmon resonance (SPR) sensors. However, the conventional SPR sensor design procedure based on the fixed parameter scanning (FPS) method is complex and time-consuming, and makes it difficult to gain the optimized design. In this paper, in order to design the optimal GH-shift-based SPR sensor, an improved differential evolution (IDE) algorithm based on chaos mapping, fitness elimination mechanism, nonlinear scale factor, and cross probability control strategy is proposed. By using such an IDE algorithm, the Ag‐ITO‐TMDCs‐graphene structure can be optimized. By the IDE algorithm, after 9 iterations, the maximum fitness value is obtained in the Ag‐ITO‐${{\rm MoS}_2}$-graphene structure, the maximum GH shift is ${16591}\lambda$, and the sensitivity is ${3.3} \times {{10}^8}\,\lambda /{\rm RIU}$. Compared with the FPS method, the GH shift is increased 192 times, and the sensitivity is increased 25,707.6 times. Compared with the DE algorithm, the number of iterations and the efficiency of the IDE algorithm are enormously improved as well. Such an algorithm provides a new, to the best of our knowledge, approach for designing a multilayer SPR sensor.

中文翻译：

---

基于改进的差分进化算法的Goos–Hänchen移位增强

Goos–Hänchen（GH）偏移是评估多层表面等离子体激元共振（SPR）传感器性能的重要方面之一。然而，基于固定参数扫描（FPS）方法的常规SPR传感器设计过程复杂且耗时，并且难以获得优化的设计。为了设计最佳的基于GH偏移的SPR传感器，提出了一种基于混沌映射，适应度消除机制，非线性比例因子和交叉概率控制策略的改进的差分进化算法。通过使用这种IDE算法，可以优化Ag-ITO-TMDCs-石墨烯的结构。通过IDE算法，经过9次迭代，可以在Ag‐ITO‐ $ {{\ rm MoS} _2} $中获得最大适应度值-石墨烯结构，最大GH位移为$ {16591} \ lambda $，灵敏度为$ {3.3} \ times {{10} ^ 8} \，\ lambda / {\ rm RIU} $。与FPS方法相比，GH偏移增加了192倍，灵敏度增加了25,707.6倍。与DE算法相比，IDE算法的迭代次数和效率也得到了极大的提高。据我们所知，这种算法为设计多层SPR传感器提供了一种新的方法。