摘要
测量的真值与被测物物理量的真值不一致,其中不可避免地含有测量误差,建立回归拟合模型是处理数据的常用方法 .通过1算例,应用附加系统参数、最小二乘配置和半参数3种模型误差补偿技术,比较其对一次函数和二次函数回归模型误差的补偿效果.结果表明,最小二乘配置得到的补偿效果最好.
The true values of measurement and the measured physical quantity are of inconsistency,which inevitably contain errors. Establishment of regression model is a common method of processing data. By an example, three kinds of model methods, including additional systematic parameters, least squares collection, and semiparametric method,are compared. We can obtain relevant conclusion through the model error compensation effect. Least squares collection method can gain better compensation effect than others.
出处
《河南科学》
2012年第9期1204-1207,共4页
Henan Science
关键词
回归模型
模型误差
补偿效果
regression model
model error
compensation effect
