Vibration Signal Denoising for Coal-Gangue Impacting the Tail Beam of Hydraulic Support Based on Improved ALO and VMD

Jialiang Dai,Lei Si,Zhongbin Wang

DOI: https://doi.org/10.1109/icemi59194.2023.10270033

2023-01-01

Abstract:Aim to the problem that the signal contains background noise when using the vibration signal of the coal-gangue impacting the tail beam of hydraulic support for coal-gangue mixed ratio identification, a noise reduction method based on the combination of improved ant-lion optimizer (ALO), variational modulus decomposition (VMD), mutual information (MI) and multiscale permutation entropy (MPE) is proposed. Firstly, an improved ALO algorithm is designed by using the chaos search strategy (CSALO) to achieve the optimization of the key parameters of VMD (VMD-CSALO). Secondly, the vibration signals of the coal-gangue impacting the tail beam of hydraulic support are decomposed by the VMD-CSALO. Finally, the denoised vibration signals are reconstructed according to MI and MPE. Then, to verify the reliability and superiority of the denoising by VMD-CSALO, a simulated experimental platform of mixed coal-gangue caving was built. Through experimental comparison with wavelet transform (WT), VMD-WT and complementary ensemble empirical mode decomposition (CEEMD)-WT methods, the results show that this paper can remove the noise more effectively, and the coal-gangue mixed ratio identification accuracy of the signal processed by this paper is significantly improved.