声表面波CO气体传感器温度误差补偿方法研究  

Study on Temperature Compensation Method in SAW CO Gas Sensors

作　　者：张朋[1] 陈明[1] 何鹏举[1] 

机构地区：[1]西北工业大学自动化学院,陕西西安710072

出　　处：《压电与声光》2010年第1期3-6,22,共5页Piezoelectrics & Acoustooptics

摘　　要：要提高声表面波(SAW)气体传感器的测量精确度,温度补偿是主要难题。目前有许多补偿方法,但其效果不佳。采用软件方法进行温度补偿的研究在国内外已成热点,但选用神经网络对声表面波气体传感器进行温度补偿罕见报道。该文以西北工业大学研发的声表面波CO气体传感器为研究对象,通过理论分析和实验,得到了声表面波CO气体传感器的温度特性曲线。提出了一种利用BP人工神经网络对声表面波CO气体传感器温度误差进行修正的新方法。计算机仿真和试验结果表明,该法能有效改善传感器的输出特性,且速度快,精度高,鲁棒性强,便于用硬件实现,具有较高的推广应用价值。It is difficult for temperature compensation of SAW CO gas sensors to improve their measure precision. In spite of many temperature compensation methods used, satisfactory result hasn＇t been achieved. Despite of the popularity of the software compensation, the neural network method isn＇t used in this sensor. In this paper, the BP neural network method is used for compensating the sensor＇s temperature error. The temperature curve of SAW CO gas sensor is obtained by the theory analysis and experiment result. The simulation result demonstrates that the SAW CO gas sensor＇s static state output characteristic can be improved effectively. Besides that, simple hardware, faster speed, high accuracy and better robustness are offered.

关 键 词：声表面波 CO气体传感器 温度误差 补偿方法 BP神经网络 

分 类 号：TN65[电子电信—电路与系统] TP212[自动化与计算机技术—检测技术与自动化装置]