Dynamic Load Identification Method Based on EEMD Decomposition

Accurate load identification in the presence of uniform noise is a crucial aspect of scientific research and engineering applications. This paper employs both the Ensemble Empirical Mode Decomposition (EEMD) algorithm and the traditional \({L}_\infty\) norm regularization method for load identification from response data. For the EEMD method, the response data is first decomposed to obtain Intrinsic Mode Functions (IMFs). These IMFs are then analyzed and quantified, using the mean and variance characteristics of the uniform noise as thresholds to filter out the IMFs that are likely to be noise components. The remaining IMFs are reconstructed to achieve denoising of the response data. Additionally, in the numerical example, data obtained from a cantilever beam finite element model with different levels of added noise are used. The monotonicity test function is applied to select the optimal regularization parameter for load identification. The relative error between the identified load and the true load is used as the evaluation criterion. The results show that the Ensemble Empirical Mode Decomposition method has higher robustness and accuracy compared to the traditional \({L}_\infty\) norm regularization method.