Temperature variation mechanism and error suppression of key parameters of phase modulator in fiber optic current sensing system

This paper focuses on the temperature reliability of phase modulators and their impact on the error of sensing systems by conducting a study on the temperature variation mechanism of key parameters and error suppression methods for phase modulators. Initially, a model correlating half-wave voltage with temperature was established and the source of polarization crosstalk was analyzed. Subsequently, tests were conducted on the half-wave voltage and polarization crosstalk of phase modulators at various temperatures to determine the relationship between temperature and these critical parameters. Then, a simulation analysis was performed to assess the impact of half-wave voltage variation and polarization crosstalk on the sensing system, revealing the pattern of system error changes. Finally, research on error suppression due to half-wave voltage changes and polarization crosstalk was carried out. A temperature variation error compensation method based on radial basis function neural networks (RBFNN) was proposed and experimentally verified. The sensor temperature, half-wave voltage, and polarization crosstalk were taken as neural network inputs, with system error as the output. The results indicate that the error in the sensing system during temperature cycling from −30 °C to + 70 °C is less than 0.04 %, which meets the 0.2 class requirements for current sensors as stipulated by national standards. This provides a reference for the application of phase modulators in fiber optic sensing and other network communication systems.