Structural parameters optimization of tapered fiber SPR sensor based on BES algorithm

Abstract

Aimed at the numerous parameters that affect the performance of fiber Surface Plasmon Resonance (SPR) sensor, which make it difficult to determine the optimal parameters by traditional mathematical modeling, on the one hand, a tapered fiber SPR sensor model is proposed to improve the detection sensitivity of sensor. On the other hand, a new intelligence algorithm with fast search speed and high search accuracy, Bald Eagle Search Algorithm (BES), is adopted to optimize the structural parameters of tapered fiber SPR sensor, including the metal film thickness (d2), sensing area length (L), taper ratio (TR), and measured refractive index (n3). At the same time, a comprehensive fitness function is defined to improve the optimization effect, which take the sensitivity, resolution and linearity of the proposed sensor model into account. In the verification process, establish three different optimization experiments, including one partition optimization, two partition optimizations and three partition optimizations. The experimental results indicate that the optimization effect of the three partition optimizations is the best, for three different optimization intervals, the obtained fitness function values are 80%, 81% and 78%, respectively. In addition, the performance of BES is compared with other five optimization algorithms, the results indicate that the best shape of SPR reflection spectrum can be obtained by BES, its average sensitivities is 2218.85 nm/RIU, which ranks first. Therefore, by three partition optimizations and BES algorithm, the optimized parameters are more representative, which make the obtained reflection spectrum based on the optimized parameters has smallest degradation. The proposed method can provide an efficient and intelligent design idea for SPR sensor, it can greatly increase the flexibility of SPR sensor design.