Application of BP Neural Network in Cross- Sensitivity Solution of Temperature and Strain of FBG Sensor

Abstract

Fiber Bragg grating (FBG) sensor has the advantages of long transmission distance and less channel occupied by data acquisition instrument, so it can be used for strain measurement of power engineering and electronic engineering and so on. FBG is sensitive to both strain and temperature, that is, both changes can change the central wavelength of the grating. In structural health monitoring, the load on the tested part is complicated, and the temperature also changes greatly in the long-term monitoring process. If the FBG sensor is fixed on the tested component, it will inevitably be affected by the combined action of temperature and strain at the same time, making the test result produce error or deviation. In this paper, the temperature compensation of FBG strain sensor is analyzed and studied. Based on the calibration data, the FBG strain sensor temperature compensation is realized by BP Neural Network Algorithm. The research results of this paper can be used as a reference for health monitoring of large-scale structures using FBG sensors.