New concept of getting data trend based on EMD and application

摘要

Aiming at the need of getting data trend in a project, the paper introduces the Empirical Mode Decomposition. According to the characteristic of real measurement data, the paper proposes a new method of getting the data trend based on Empirical Mode Decomposition. The new concept validates the noise distributing rule of measure data and reflects the characteristic of Empirical Mode Decomposition. Compared with the polynomial filter, the new algorithm keeps the integrality of data information, effectively reduces the effect of noise, improves the reliability of data information, and gets the data trend. Through simulation result, the new concept is proved to be valid and applicable.