从“随机过程”的角度认识测量误差及其分类

The Cognition of Measurement Error and Classification from the Perspective of “Stochastic Process”

测量误差作为"随机过程"具有统计特性,通过自相关函数、功率谱密度等特性参量对"随机过程"中各因素引起的误差分量特点进行分析,并得出结论:经典误差理论中系统误差和随机误差之分应该即是对此"随机过程"的实用分类,其实质是"自相关性"的两种极限情况,系统误差是自相关系数等于1的完全正相关,随机误差是自相关系数等于0的完全不相关。

To points out that the measurement error has statistical characteristic when as a ＂stochastic process＂, it has been done that to analyze the error component＇s characteristic caused by different factor in ＂stochastic process＂ with the characteristic parameters such as self-correlation function and power spectrum density, and to come to the conclusion that the systematic error and random error in classical error theory should be the practical classification of this ＂stochastic process＂. The essences of systematic error and random error are two limit cases of ＂self-correlation＂. The systematic error is a completely positive correlation with the self-correlation coefficient equal to 1, and the random error is a completely none correlation with the self-correlation coefficient equal to 0.