摘要
应用BP神经网络技术抑制温度对光纤光栅压力传感器的干扰,从而提高了压力传感器的选择性.以聚合物封装的光纤光栅传感器为例,当温度从19℃变化到75℃时,光纤光栅布拉格波长偏移量是11.315 nm,由此导致传感器输出的引用误差为1 915%.经神经网络融合处理后,其值降为2%,实现了对压力较为准确的识别.实验结果表明,该方法具有实际应用前景.
To increase the selectivity of fiber Bragg gratings pressure sensor, a method is proposed by applying the back propagation(BP) neural network which can restrain the temperature disturbance to pressure sensor. Take the fiber Bragg grating sensor that the polymer packages as example , when temperature changes from 19℃ to 75℃, the optical fiber gratifig Bragg wavelength excursion is 11. 315 nm. So the error of the sensor exports is 1 915%. With the aid of BP neural network ,it can be decreased to 2% . Consequently, pressure is detected accurately. The method provides an effective approach for improving selectivity of fiber Bragg gratings pressure sensor and exhibits practical prospect.
出处
《传感技术学报》
CAS
CSCD
北大核心
2007年第7期1531-1534,共4页
Chinese Journal of Sensors and Actuators
基金
中国石油天然气集团公司应用基础研究项目资助(20050719)
陕西省教育厅产业化培育项目资助(05JC23)
西安市科技局信息技术专向项目资助(2X05041)
西安石油大学科技创新基金资助(2004-24)
国家863项目资助(2006AA06Z210)
国家自然科学基金项目资助(60654001)