基于光电调制式的多点多气体光纤传感网络

摘要

In order to explore a long distance fiber sensing network in the multi-gas environment, a multi-point multi-gas sensing optical fiber network based on photoelectric modulation was proposed and demonstrated.This system combines the advan-tages of electric sensing with the superiority of long haul optical fiber network.The back propagation neural network platform(BP network)was employed to eliminate the cross interference between the various gas sensors.The sensing network covers more than 25 km distance and multiple channels.In the experiment,the gas chamber was filled with mixed H 2,CO2,and CH4,and the gas concentrations detected quantitatively with the elimination of cross interference.The sensitivity of measurement was about 100ppm.The temperature impact analysis was carried out to further reduce the system concentration error,and the deviation can be reduced.The overall system error is about 4ppm.Its network test effect is good.%针对混合气体环境下长距离传感网络构建需求,提出并实现了一种基于光电调制式的多气体一体化的光纤传感网络结构.该系统结合了电学式探测和光纤组网传输的双重优势,搭建基于误差前向反馈神经网络算法(BP神经网络)的软件平台以消除交叉干扰,实现了25 km以上多点多气体光纤传感网络.实验中成功实现对H2、CH4、CO23种混合气体同时定量浓度检测和交叉干扰消除,测量灵敏度100 ppm左右,并进行了温度影响分析,进一步降低系统浓度误差,整体系统误差为4 ppm左右,其网络测试效果良好.