摘要
提出了一种基于递归神经网络的热电偶测温滞后的动态补偿方法,利用神经网络良好的非线性映射能力,建立传感器的动态逆模型,实现对传感器的动态补偿。实验结果表明:检测信号经过动态补偿后,能够克服传感器的测量滞后,达到稳态的时间从补偿前的26 s缩短到大约5 s,传感器的动态性能得到较大的提高。
The method of dynamic compensation based on recurrent neural network to correct the thermocouple temperature sensor＇ s error is introduced and the dynamic Compensation of temperature sensor is realized by building sensor＇ s dynamic inverse modle. The experimental results show that the measurement lagging of sensor can be overcome, the time reaching to stable performance can be decreased from 26 s to 5 s and the dynamic performance can be improved after dynamic compensation .
出处
《传感器与微系统》
CSCD
北大核心
2006年第11期38-40,共3页
Transducer and Microsystem Technologies
关键词
温度传感器
递归神经网络
动态补偿
temperature sensor
recurrent neural network
dynamic compensation
