Owing to the difference in the response of photodetectors (PDs) and the difference in losses between different optical paths, the conventional three-wavelength demodulation scheme requires relatively complex early calibrations to maintain sensitivity. Additionally, a multi-optical path structure requires multiple PDs for synchronous detection and multichannel data acquisition, which increases the volume and cost of the system. Moreover, this scheme requires an upper computer to achieve sound-signal demodulation, thus significantly restricting on-site sound signal detection. To further optimize the feedback control logic and simplify the demodulation-system structure, this study investigates a three-wavelength adaptive intensity-demodulation technology based on a fast-tunable modulated grating Y-branch (MG-Y) laser. The demodulation system can flexibly adjust the operating-wavelength interval based on the initial cavity length of the sensor to flexibly match fiber-optic acoustic sensors with different initial cavity lengths. The storage, transmission, and feedback logic judgment of the data are implemented using field-programmable gate array (FPGA) modules, which can demodulate complex acoustic signals without requiring complex structures such as upper computers.

The three-wavelength adaptive intensity-demodulation scheme was classified into parameter-calibration and real-time detection stages. In the parameter-calibration stage, the phase difference caused by the corresponding optical-path difference between the three operating wavelengths in the same Fabry-Perot (F-P) cavity was 2π/3. Three operating wavelengths with a 2π/3 phase difference were determined based on the initial cavity length of the sensor. In the absence of an acoustic signal, the reflected light intensity and direct flow rate of the three operating wavelengths were sequentially detected, and the threshold range of the sensitive area was calculated. In the real-time detection stage, the reflected-light-intensity data obtained in real time were used to determine whether the current operating wavelength was within the sensitive area and capable of wavelength-adaptive adjustment. Suitable wavelengths were output among the three working wavelengths to ensure that the acoustic sensing system maintained a high detection sensitivity. The detection sensitivity of the system remains at 0.866 times (or higher) the highest sensitivity.

Using the MG-Y laser for full-spectrum scanning, we obtained relationship between interference spectrum and interference intensity at different output wavelengths. The free spectral range is 4.92 nm and the initial cavity length of the sensor is 243.23 μm (Fig. 4). Subsequently, the switching of the work points was monitored (Fig. 5). The sound pressure was calibrated using an electrical microphone, the speaker was controlled to emit sound signals ranging from 1 to 20 kHz, and the frequency response of the sensor was tested (Fig. 7). The frequency of the signal emitted by the speaker was controlled to remain constant, the sound pressure was varied, and the demodulated waveforms were compared under different sound pressures. The demodulated waveforms have the same frequency and the amplitude is proportional to the sound pressure. The speaker was controlled to emit a signal with a constant sound pressure. Subsequently, the frequency was changed and the demodulated waveforms under different sound pressures were compared. The demodulated waveform frequency is directly proportional to the sound signal frequency (Fig. 8). The speaker was controlled to play audio. Subsequently, the audio signals received by the electrical microphone and diaphragm sensor were recorded and compared. Based on observation, the demodulation system can demodulate multifrequency mixed sound-pressure signals (Fig. 9). The demodulation of the system was continuously tested for a long time and 70 h of sample detection was recorded. The amplitude of the demodulated signal was recorded as a function of time. Calculations show that the minimum amplitude was approximately 0.868 times the maximum amplitude, whereas the minimum sensitivity was approximately 0.868 times the maximum sensitivity. Thus, the principle was validated (Fig. 10).

In this study, a three-wavelength adaptive intensity demodulation technology and an acoustic sensing system based on a modulated grating Y-branch laser are proposed to solve the complex structure and feedback control logic of conventional fiber optic F-P acoustic sensing systems. The proposed three-wavelength adaptive-intensity demodulation scheme utilizes a single widely tunable MG-Y laser to achieve flexible control of the operating wavelength. Experimental results show that the three-wavelength adaptive strength-demodulation system presents high resistance to environmental interference, favorable demodulation performance, high stability, and the ability to demodulate multifrequency mixed sound-pressure signals. The proposed demodulation scheme can be implemented easily using microcontrollers or FPGAs without requiring a host computer, which is conducive to the development of low-cost, miniaturized systems. It may facilitate the effective detection of sound signals in harsh environments, such as those with high temperature, high pressure, and strong electromagnetic interference.