高斯过程回归在光纤光栅隧道火灾测温系统中的应用

Gaussian Process Regression Applied to Fiber Bragg Grating Tunnel Fire Temperature Measurement System

光纤光栅测温技术以其现场无源、灵敏稳定的特点,非常适用于隧道环境的火灾监测。光栅测温系统目前最常用的测温模型是通过二次曲线拟合得到,其形态固定,存在一定的误差。文中提出基于高斯过程回归方法的测温模型,能够更加恰当地描述光栅中心波长随温度变化的规律,具有更高的准确度。

Fire is one of the high risks of highway tunnels. Fiber Bragg grating temperature measure-ment technology has the characteristics of passive detection, sensitive and stable. It is suitable for tunnel fire monitoring. At present the most commonly used temperature measurement model is ob-tained by quadratic curve fitting. It has fixed form and a certain error. A temperature measurement model is presented in the paper based on Gaussian process regression method. It could describe the change rule of grating wavelength with the temperature more aptly, and has higher accuracy.