从坏值概率分析基础上提出了动态测量中传感器数据的坏值剔除方法,用可靠的数据代替剔除的坏值,为后续的传感器数据处理保证了数据的合理性,如传感器的非线性拟合。并通过对动态测量中某霍尔位移传感器位移- 电压数据进行了坏值剔除和三次多项式非线性拟合,表明动态测量中传感器数据的非线性处理方法。 关键词: 非线性; 坏值剔除; 动态测量; 曲线拟合 Abstract : After analysis the probability of abnormal data , a method erasing the abnormal data of non - linear sensor in dynamic measurement is given.Based on the method ,the abnormal data is replaced by reliable one ,and curve fitting method is given to deal with the processed data. The curve of a Hall2displacement sensor is obtained by polynomial fitting. The method of data2erasing and curve2fitting can be applied in processing the data of sensor in dynamic measurement . Key words : non2linearity ; abnormal data erasing ; dynamic measurement ; curve fitting