摘要
离子传感器是环境水质监测的关键技术之一。由于传感器制备技术的局限性以及应用环境的复杂性,严重影响了传感器的测量精度以及可靠性,难以用于在线检测。为了给水质在线检测提供有效的分析手段,以铵离子传感器为背景,在深入研究传感器响应特性后,根据实验数据,使用支持向量机回归建立传感器响应的离线模型;进一步使用在线支持向量机回归建立传感器的漂移校正模型。实验结果表明:建立的模型能较好地拟合响应过程曲线,有效地补偿了传感器的漂移。
Ion selectivity sensors( ISEs) are one of the key technologies in environment and water quality monitoring. Due to the limitations of the sensor's fabrication technology and the complexity of the application environment,it is very difficult to use ISEs for online monitoring on water environment. In order to provide an effective analysis method for the online measurement of water quality,the response characteristics of ammonium ISE is investigated,and SVR( support vector machine) is introduced to model ammonium ISE off-line. Furthermore,the drift calibration model of ammonium ISE is built by using online SVR. The experimental results show that the approach can fit the response curve of ammonium ISE accurately and compensate the sensor drift effectively.
出处
《仪表技术》
2015年第8期47-51,共5页
Instrumentation Technology
关键词
离子传感器
支持向量机
在线支持向量机
传感器漂移
校正
ion selectivity electrode(ISE)
support vector regression(SVR)
online support vector regression
sensor drift
calibration
