一种基于EEMD的过程数据混合去噪方法

Hybrid process data denoising method based on EEMD

针对化工过程数据的特点,提出一种基于集成经验模式分解(EEMD)滤波的过程数据混合去噪方法,以新秩一阶差分法抑制数据粗差干扰,以EEMD分解抑制脉冲干扰,分层滤波消除噪声成分。与传统的滤波方法相比,基于EEMD的混合滤波方法无须预先确定滤波器参数,是一种完全的数据驱动型方法,具有较好的自适应能力。仿真实验结果表明,对过程数据的滤波预处理可以增强对异常突变数据的检测处理,提高故障检测效果。

In view of the features of chemical process data,this paper presented a novel integrated EEMD preprocessing met-hod based on the principles of EMD denoising and first order differential.This method designed a first order differential method using new rank as the pre-filter process unit to reduce the effects of gross errors,and used a denoising scheme based on EEMD method to suppress pulse interference and remove white noise from the signal.Compared with traditional filtering,the hybrid EEMD filtering method did not need to define the coefficients of filter,so it was fully data-driven and adaptive.The simulation and experimental results demonstrate the effectiveness of this method in gross error elimination and fault detection.