Abstract

The performance of Gaussian process regression for temperature estimation using fiber Bragg grating sensors is investigated. Using experiment- and simulation-based training, the estimated temperature uncertainty (standard deviation) and offset are analyzed versus different measurement parameters.