基于差分进化算法的光纤布拉格光栅 横向局部应力与温度分离技术 = Simultaneous measurement of stress and temperature of fiber Bragg grating under local transverse force using differential evolution

• Main Authors: 陶继平 Tao Jiping, 刘暾东 Liu Tundong, 傅晓立 Fu Xiaoli, 陈静 Chen Jing, 江灏 Jiang Hao
• Other Authors: School of Electrical and Electronic Engineering
• Format: Article
• Language: Chinese
• Published: 2018
• Subjects: 交叉敏感 Cross-sensitivity; DRNTU::Engineering::Electrical and electronic engineering; 光纤布拉格光栅 Fiber Bragg Grating
• Online Access: https://hdl.handle.net/10356/82752; http://hdl.handle.net/10220/46649
• Institution: Nanyang Technological University

鉴于横向应力和温度对光纤布拉格光栅的交感耦合呈现非线性关系，将横向局部应力与温度的交感问题转化为非线性系统参数辨识问题: 首先根据光纤布拉格光栅横向应力与温度的传感原理构造一个理论意义上的反射光谱，然后利用理论光谱和采样光谱之间的差异度建立系统参数辨识模型，通过差分进化算法对该优化辨识模型进行求解，最终实现横向局部应力与温度的同时测量。实验和仿真结果表明，本文提出的方法具有较高的辨识精度，得到的横向应力误差绝对值一般小于4．0 ×10 － 3N，温度误差绝对值一般小于3． 0 × 10 － 3℃，传感器对温度的灵敏度为11． 7 pm/℃。该方法能够解决光纤布拉格光栅横向局部应力和温度的交叉敏感问题，提升光纤布拉格光栅传感网络在实际工程中的传感精度。Considering that fiber Bragg grating ( FBG) sensors are sensitive to both transverse stress and temperature simultaneously, which leads to the difficulty in the independent measurement of these two measurands, we transform this problem into nonlinear system parameter identification with identification parameters of temperature and stress. A theoretical reflection spectrum is first constructed based on the sensing principle of FBG transverse stress and temperature. Then a novel identification model is established based on the discrepancy between actual measured spectrum and the constructed spectrum. The differential evolution (DE) algorithm, a simple yet powerful evolutionary algorithm, is applied to solve the proposed model and obtain the appropriate values of temperature and stress simultaneously. Experiment and simulation results show that the measurement error of transverse stress is typically smaller than 4.0 × 10 － 3N and the measurement error of temperature is smaller than 3. 0 × 10 － 3℃, the temperature sensitivity of the sensor is 11.7 pm/℃. The result indicates that the proposed approach is effective in discriminating the cross-sensitivity of FBGs, while maintaining high measurement accuracy.