Fiber optic current sensor temperature compensation through RBF neural network

摘要

The fiber optic current sensor (FOCS) is susceptible to external temperature in actual operation, which will lead to its accuracy deviation, even malfunction. In order to improve the temperature stability of FOCS's ratio error, a temperature compensation method based on RBF neural network is established by taking the temperature as input and the ratio error as output to the network. Compared with BP neural network, the simulation results show that the temperature compensation model based on RBF neural network has better accuracy whose prediction error is less than 3%. At the same time, the experimental results show that the drift deviation of ratio error can remain as low as ±0.1% in the range of-40 °C to 70 °C, and the 0.2S-level accuracy of GBT20840.8 standard can be achieved.