分布式光纤喇曼测温系统采用了线性逼近的方法进行温度解调, 并通过灵敏度分析法获取了相应的线性解调参数。为了提高系统的精度, 在信号累加平均去噪的基础上, 采用了卡尔曼滤波的方法, 通过对前一时刻的估计值与当前测量值的递推, 实现了对存在测量噪声的温度信号的最优化估计, 完成了对现场采集信号的实时更新和处理。实验验证了该滤波方案的可行性, 随着外界环境温度的变化, 测温系统能够得到合理的温度曲线, 测量误差小于１℃。系统响应性良好,性能稳定, 能够适应复杂的环境变化。

The distributed optical fiber Raman temperature measurement system adopt the linear approximation for temperature demodulation and obtained the relating demodulation parameters through sensitivity analysis. In order to improve the measurement precision, the Kalman filter is applied in addition to the signal cumulative average denoising technique. The optimal estimation for the temperature signal with disturbance is achieved through the recursion of the previous data estimation and the current measurement data. The real-time update and processing of the measurement data is realized. The feasibility of the filtering method is verified by the experiment which suggested that the system can obtain reasonable temperature curves with measurement error less than 1℃ following the variance of the environment temperature outside. The system has good performance of stability and responsibility, also can adapt to the complex environmental changes.