A In-situ Calibration Method for Fiber Grating Strain Sensors Based on Ga-bp Neural Network
The center wavelength offset affects the accuracy of the results of fiber grating strain sensors,and an in-situ calibration method for fiber grating strain sensors based on GA-BP neural network is proposed.The wavelength drift of the fiber grating sensing center is determined and a BP neural network model is built.GA algorithm is used to improve the connection weight and threshold of the BP neural network.The center wavelength offset is used as the input parameter of the GA-BP neural network.The central wavelength,peak width,reflection peak intensity,nonlinear error and temperature are used as inputs to the GA-BP neural network.The central wavelength,peak width,reflection peak intensity,nonlinear error and temperature offset are removed by the GA-BP neural network to obtain new input parameter values.Using these values,the parameters of the fibergrating strain sensor are readjusted to achieve in-situ calibration of the fiber grating strain sensor.The experimental results show that after using the proposed method for in-situ calibration of fiber grating strain sensors,the center wavelength,peak width,reflection peak intensity,nonlinear error and temperature variation obtained are mostly consistent with the ideal variation.The strain value is highly consistent with the calibration value,indicating that the calibration effect of this method is good.
NETL
NSTL
万方数据
