A Joint Denoising Method Based on Adaptive Large Neighborhood Search, Ensemble Empirical Mode Decomposition, and Affine Projection Algorithm
Cheng Zhang, Yuhui Gao, Jingxi Wang, Lingqi Meng, Zhefei Wang, Ran He · 2025
With the widespread application of composite insulators in power systems, ensuring their internal structural integrity is crucial. To address the issue of noise interference in ultrasonic detection signals of composite insulators, this paper proposes a joint denoising method based on Adaptive Large Neighborhood Search (ALNS) optimization, Ensemble Empirical Mode Decomposition (EEMD), and the Affine Projection Algorithm (APA). ALNS is used to optimize the parameters of EEMD decomposition, effectively solving the mode mixing problem in traditional decomposition methods. EEMD leverages its multiple white noise decomposition properties to achieve efficient signal decomposition and adaptive threshold denoising. Finally, APA filtering further enhances signal quality, significantly improving the signal-to-noise ratio (SNR) and reducing the mean squared error (MSE). Simulation and experimental results demonstrate that the proposed method performs excellently in complex environments, providing a more precise and robust signal optimization solution for ultrasonic detection.