一种基于EMD的系统误差分离方法

摘要

现有的测量误差分离方法大多需要预先确定误差的变化特征和基函数,而基函数的选取没有明确的标准.该文提出了一种采用自适应分解的经验模态法对周期较大的系统误差进行提取和自相关分析与经验模态相结合的分解法对周期较小的系统误差进行提取的方法.通过实际仿真结果表明,自相关分析与经验模态分解法不需要确定误差特征就能够成功分离系统误差与随机误差,达到了预期效果.%At present, the variation feature and primary function must be known firstly in available measurement error separation methods, and the choice of primary function is random.Two methods, extracting the long-period systematic errors by using EMD of adaptive separation and extracting the short-period systematic errors by using the autocorrelation analysis and EMD,were proposed in the article.The simulation results showed that the method of autocorrelation analysis and EMD could separate the systematic errors and random errors, and reached the prospective effectiveness.