一种基于优化阈值函数的整体强化分解模型改进降噪方法

An Improved EEMD De-noising Method Based on Optimized Threshold Function

以提高降噪效果为目的,提出了一种基于白噪声分解特性的EEMD优化阈值降噪方法。避免了小波分解时选择合适小波基函数的困难,具有自适应性,同时,也可以有效地避免频率混叠问题。方法从能量密度和对应平均周期乘积的变化率出发,提出了"跳变点"和"奇异点"的选择原则,并利用优化阈值函数对"跳变点"和"奇异点"对应的IMF分量进行量化处理,然后通过对处理后的IMF分量重构得到降噪信号。最后用仿真信号进行试验,证明本方法的有效性和优越性。

An EEMD optimized threshold de-noising method was proposed based on the decomposed characteristics of white noise for the purpose of de-noising. The method is adaptive, free from the choice of wavelet base and the determination of the number of decomposition order, and at the same time it can effectively suppress the phenomenon of mode mixing. The choice principle of mutation point and bizarre point was proposed start with the change rate of the product of the energy of IMF component and its corresponding averaged period. Then the method quantifies the mutation point IMF and bizarre point IMF by the way of optimized threshold de-noising function proposed. It can get the de-noising signal by summing the quantified IMFs. At last, the experiment to the simulation based signal proves the valid and superiority of this method.