Optical resonators are particularly suitable for anaemia state detection due to small sample size, real-time detection and low power consumption. However, the quality factor and sensitivity of resonant cavity sensors are mutually constrained, so the single objective optimization algorithms proposed so far can only achieve a single optimization of sensitivity or quality factor, which limits chip performance. This article presents a multi-objective optimization design of the runway ring resonant cavity sensor based on the BP-NSGA II algorithm, which has achieved good performance improvement. The simultaneous incorporation of quality factor and sensitivity provides access to key structural design parameters to overcome the limitations of sensitivity and quality factor constraints on each other, thereby improving sensor performance. The results show that the optimized structure has a sensitivity of 439 nm/RIU, a quality factor of 938, a relative error of 1.37% and 3.9% for the sensitivity and quality factor, respectively, a linearity of 0.00004636% for the sensitivity, and a training time of 3 min. The method has the advantages of small errors and short learning time. A new optimization method is provided for the design of the resonant cavity of a runway ring.

光学谐振腔由于样本量小、检测实时且功耗低，特别适合贫血状态检测。然而，谐振腔传感器的品质因数和灵敏度是相互制约的，因此目前提出的单目标优化算法只能实现灵敏度或品质因数的单一优化，限制了芯片性能。本文提出了基于BP-NSGA II算法的跑道环谐振腔传感器的多目标优化设计，取得了良好的性能提升。品质因数和灵敏度的同时结合提供了对关键结构设计参数的访问，以克服灵敏度和品质因数相互约束的限制，从而提高传感器性能。结果表明，优化后的结构灵敏度为439 nm/RIU，品质因数为938 ，灵敏度和品质因数的相对误差分别为1.37%和3.9% ，灵敏度线性度为0.00004636% ，训练时间3 分钟。该方法具有误差小、学习时间短的优点。为跑道环谐振腔的设计提供了一种新的优化方法。