Objective Nowadays, the demodulation precision, speed and stability of existing fiber grating demodulation systems have been greatly improved. However, most FBG demodulation systems are complex in structure, low in integration, poor in engineering adaptability, and the embedded peaking algorithm is complex, so the demodulation needs to be completed by the host computer, which is poor in portability and high in cost. Therefore, the research goal of this paper is to design an FBG demodulation system which is simple in structure, high in integration and can satisfy the large-scale engineering application of FBG.

Methods In view of the above problems in the current fiber Bragg grating demodulation system, a FPGA high-precision FBG-aware demodulation system based on adaptive peak finding algorithm is designed. Firstly, the sensing mechanism of FBG is studied, and a photoelectric detection module is designed to convert and amplify the weak reflection signal of FBG into a voltage signal. Based on the corresponding relationship between the synchronous trigger signal output by the tunable laser and the scanning wavelength, the FPGA microcontroller realizes the efficient acquisition of the converted FBG reflection signal. Secondly, the Kalman moving mean hybrid filtering algorithm was introduced into the FPGA microcontroller to denoise the original spectral reflection signal, the peak judgment algorithm was used to realize the adaptive multi-peak region discrimination, and the double centroid algorithm was combined to complete the fast and high-precision demodulation of the center wavelength of multiple FBG reflected light.

Results and Discussions A FPGA high-precision FBG sensing demodulation system based on adaptive peak finding algorithm is built (Fig.5). The Kalman moving mean hybrid filter is used in the FPGA to denoise the collected original spectral data and effectively correct the spectral shape (Fig.6). The experimental results show that under the same conditions, the proposed algorithm has faster demodulation speed than the traditional algorithm, while maintaining better demodulation precision and stability, and consumes less underlying hardware computing resources (Fig.7, Tab.1). When the temperature is 37 ℃, 37.5 ℃ and 38 ℃ respectively, the central wavelength value of the demodulated system in this paper fluctuates within the range of ±3 pm compared with the standard value (Fig.8). When the temperature rises from 25 ℃ to 35 ℃, the correlation coefficient between temperature and wavelength change is above 0.998 5, and the demodulation results of the system have a good linear relationship (Fig.9). Based on the linear experiment, two cooling tests and one heating test were repeated. The temperature measured by the three demodulation experiments is in good agreement with the actual measured temperature without obvious hysteresis and drift (Fig.10), and the repeatability of the demodulation results of the system is good. The multi-grating demodulation of the system is tested experimentally, which still shows good linearity and demodulation accuracy (Fig.11, Tab.2).

Conclusion  Through theoretical and experimental research, combined with tunable laser, a FPGA high-precision FBG sensing demodulation system based on adaptive peak finding algorithm is designed and built. The system is based on FPGA platform, the photoelectric detection module is designed, and the Kalman-moving mean hybrid filter algorithm is embedded to reduce the noise of the spectral signal, which effectively reduces the influence of noise disturbance on the peak finding accuracy. The peak judgment algorithm and the double centroid algorithm are used to realize the fast and high-precision demodulation of the peak center wavelength of multiple FBG, which effectively improves the demodulation accuracy. To some extent, it solves the shortcomings of the current fiber Bragg grating demodulation system, which is difficult to deploy in the demodulation equipment and requires the help of the host computer for demodulation. The temperature measurement experiment shows that the demodulation accuracy of the central wavelength of the FBG temperature sensor can reach ±3 pm, the fitting linearity of temperature and wavelength change is above 0.998 5, and the system shows excellent stability and repeatability. It provides a useful reference for the portability, integration and engineering of fiber Bragg grating demodulation system.