论文部分内容阅读
针对光寻址电位传感器(LAPS,light addressable potentiometric sensor)测量精度易受温度影响的问题,提出一种基于支持向量机(SVM)的LAPS温度补偿方法。根据在多温度条件下LAPS传感器对缓冲液pH的实测数据,利用SVM建立LAPS温度补偿模型,通过核函数把LAPS输出饱和光电流与温度间的非线性关系映射到高维特征空间,在高维空间中用线性回归实现该映射的非线性处理,对LAPS的温度特性进行非线性逼近,补偿温度对LAPS输出的影响。实验结果表明,在(20±1)、(20±2)和(20±5)℃温度变化情况下,经过温度补偿后LAPS系统输出pH的标准偏差分别为补偿前的1/3、1/4和1/6,均小于0.05pH,LAPS温度补偿算法可以明显补偿温度影响,提高LAPS的测量精度。
Aiming at the problem that the measurement accuracy of light addressable potentiometric sensor (LAPS) is easily affected by temperature, a method of LAPS temperature compensation based on support vector machine (SVM) is proposed. According to the measured data of LAPS sensor’s buffer pH under multi-temperature conditions, the temperature compensation model of LAPS is established by using SVM, and the non-linear relationship between LAPS output saturation current and temperature is mapped to high-dimensional feature space by kernel function. Linear mapping of space is used to realize the non-linear processing of this mapping. Nonlinear approximation of LAPS temperature characteristics is performed to compensate for the influence of temperature on LAPS output. The experimental results show that the standard deviation of the output pH of the LAPS system after temperature compensation under the temperature changes of (20 ± 1), (20 ± 2) and (20 ± 5) ℃ is 1/3 and 1 / 4 and 1/6, both less than 0.05pH, LAPS temperature compensation algorithm can significantly compensate for temperature effects, improve LAPS measurement accuracy.