High-precision pressure sensor based on Fiber Ring Resonator empowered by machine learning

In this study, the use of Fiber Ring Resonators (FRR) as highly sensitive pressure sensors is investigated. Enhanced system stability and sensitivity are achieved by employing the Pound-Drever-Hall (PDH) frequency-locking technique. The shift in the resonance of the FRR due to external pressure variations is used to measure the applied pressure with high precision. Conventional pressure sensing methods rely on monitoring the PDH error signal generated by the lock-in amplifier. However, the output of the PDH system is not a directly readable value but rather a composite signal influenced by the PI controller and the characteristics of the system hardware. In this work, a machine learning-based data processing method is introduced and demonstrated in the FRR system to achieve high-precision pressure measurement. This paper presents the theoretical background, experimental setup, and results demonstrating the feasibility and advantages of this approach.