摘要
由于温度、湿度等环境参数对半导体气体传感器的测量精度有着较大的影响,对此提出了基于神经网络方法的智能气体传感器设计方法,利用神经网络对气体传感器的各种环境参数进行数据融合,以获取准确的被测气体浓度值。实验证明,该方法可以有效地提高气体传感器的测量精度,使气体传感器输出的满量程误差在±0 8%以内。通过该神经网络融合逆模型,还可以方便地根据环境参数的变化实现对气体传感器输出的补偿,从而实现传感器的智能化。
The effect of change in environment conditions, such as temperature and humidity, on the gas sensor is great. A scheme of an intelligent gas sensor using an artificial neural network (ANN) is proposed. In order to improve the accuracy of gas sensor, the inverse modeling based ANN is used to estimate the applied gas concentration. It is revealed from the simulation studies that this inverse model can provide correct concentration readout within ±0.8% full-scale error. When there is a change in ambient temperature and humidity, the inverse model can compensates for this change automatically. Consequently, the intelligent sensor is achieved.
出处
《计量学报》
CSCD
北大核心
2004年第4期380-383,共4页
Acta Metrologica Sinica
基金
广东省自然科学基金(32030)
关键词
计量学
气体传感器
神经网络
环境参数
智能传感器
Metrology
Gas sensor
Artificial neural networks
Ambient parameters
Intelligent sensor
