A novel algorithm for highly sensitive and rapid hotspot detection in HTS magnets using quasi-continuous fiber Bragg gratings

Most optical fiber sensors based on Brillouin/Rayleigh scattering, and discrete fiber Bragg gratings (FBGs) cannot offer the speed and distributed sensing with operation in a cryogenic environment < 80 K necessary for superconducting magnet protection. An ultra-long fiber Bragg grating (ULFBG) comprised of quasi-continuous FBGs has the potential to rapidly detect the occurrence of a thermal hot spot in high-temperature superconducting magnets. However, processing the data for detection purpose is challenging. This is because the spectrum of an ULFBG is complex, and the hotspot induced spectral changes are difficult to identify. In this paper, we present the novel Spectral Intensity Change (SIC) algorithm to 1) extract the region of interest within the spectrum based on unique change patterns, 2) identify the direction of temperature change, and 3) process the isolated spectral data to signal hotspot events. To demonstrate the effectiveness of the algorithm, three ULFBGs with sensing lengths of 0.2, 0.5 and 1.9 m are used to detect events of temperature variations. Using the algorithm, ULFBG sensors not only can identify a hot spot event rapidly and accurately but also achieve a significantly enhanced signal to noise ratio (SNR) at 80 K. The SIC algorithm has the potential to be implemented for real-time hotspot and quench detection in superconducting magnets.