摘要
气体传感器与普通传感器特性存在很大差别,比较突出的是具有严重的滞后。由于滞后的存在,在线检测时,传感器响应大多数无法达到稳态值。这些滞后包括线性与非线性两种。针对线性滞后,建立了一种一阶线性滞后模型,利用瞬态过程的中间数据进行回归,实现了稳态值的最佳估计,并通过实验证实了该模型具有较高的准确度。针对非线性滞后,提出了一种基于移位回归的算法,该算法可以有效地消除非线性滞后,并且具有高度自适应能力。此外,对移位回归算法进行了优化。该方法不仅适用于电子鼻系统,也适用于各种气体传感器信号检测的应用场合。
There is big difference between the gas sensor and the ordinarysensor,and the more prominent is serious lag. Due to the presence of hysteresis,in the case of the line detection,the sensor can not achieve steady-state value in response to the majority. These include linear hysteresis and nonlinear. For linear lag,a linear first-order lag model was established. Using the intermediate data transients regression,the best estimate of the steady-state value was achieved,and experiments confirmed that the model has high accuracy. Nonlinear hysteresis,it proposes an algorithm based on the shift regression,the algorithm can effectively eliminate the nonlinear hysteresis,and has high adaptive capacity. In addition,the regression algorithm is optimized. This method is applicable not only to electronic nose system,but also for a variety of gas sensor signal detection applications.
出处
《仪表技术与传感器》
CSCD
北大核心
2016年第10期24-27,共4页
Instrument Technique and Sensor
基金
国家重大科学仪器设备开发专项项目(2012YQ060165)
关键词
电子鼻
两种滞后
响应模型
移位回归
electronic nose
two types of hysteresis
response model
shift regress
