基于最小二乘支持向量机的车用铂电阻温度传感器非线性校正  

Nonlinear correction of Pt resistance temperature transduce in vehicles based on least square support vector machine

作　　者：蒋寿生[1,2] 鄂加强[1] 刘云卿[1] 龚金科[1] 

机构地区：[1]湖南大学机械与汽车工程学院,湖南长沙410082 [2]邵阳学院能源与动力工程系,湖南邵阳422000

出　　处：《湖南农业大学学报（自然科学版）》2008年第5期613-616,共4页Journal of Hunan Agricultural University(Natural Sciences)

基　　金：国家自然科学基金项目(50571069)

摘　　要：依据铂电阻温度传感器样本,提出了一种采用最小二乘支持向量机辨识传感器逆模特征的校正铂电阻温度传感器非线性误差的原理和方法.该方法不需逆模型函数形式的先验知识,能够保证找到的极值解就是局最优解,具有较好的泛化能力.实例应用表明,其检测精度可达到0.1℃,铂电阻温度传感器非线性校正效果好.According to the sample of Pt resistance temperature transduce, a method of correcting nonlinear error of transduce, which could identify the contrary model＇s characteristic of the transduce correctly, was given based on the least square support vector machine. The method didn＇t require any prior knowledge about contrary model function and could ensure that extreme solution was optimal and produced a strong commonality. The application results reveal that transduce error is less than 0.1 ℃, nonlinear correction of Pt resistance temperature transduce is of good result.

关 键 词：最小二乘支持向量机 铂电阻温度传感器 非线性校正 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]