摘要
介绍了利用BP神经网络对传统热导气体分析仪改善的新方法。通过大量的样本数据训练构建了双输入早输出网络模型,通过单片机来实现整个控制处理过程。结果表明:仪器测量精度有了很大提高,采用改进的算法可以实现传感器高精度温度补偿。
A new method of improving thermal conductance gas analyzer was introduced, which adopted BP neural network. A double imput single output net model was constructed by training a mass of sample data. The control and process was realized by the monolithic integrated circuit. The results shows that the precision of analyzer is improved greatly and high precision temperature compensation of sensors can be realized with this improved arithmetic.
出处
《仪表技术与传感器》
CSCD
北大核心
2007年第12期62-63,73,共3页
Instrument Technique and Sensor
关键词
温度补偿
BP神经网络
热导
： temperature compensation
BP neural network
thermal conductance
