Research on vessel noise signal compression sensing algorithm based on Particle Swarm Optimization and Variational Mode Decomposition

Abstract

Due to the complexity of marine environmental factors and strong background noise, the collected vessel signal becomes intricate and challenging to transmit and reconstruct using conventional parameter estimation methods. Additionally, identifying vessel targets becomes difficult. In order to detect the vessel target and obtain the corresponding orientation information, it is necessary to eliminate the interference of noise as much as possible and reconstruct the original signal more accurately. In this work, we propose an optimized signal denoising method PSO-VMD (Particle Swarm Optimization–Variational Mode Decomposition). By conducting simulations using experimental data and evaluating the model, we demonstrate the superior denoising capabilities of this method. Then we apply this method to denoise the real vessel sound signal and then compress and reconstruct the signal using compressed sensing. The experimental results show that the signal processed by PSO + VMD method has the maximum SNR and the minimum RMSE and MAE. What’s more, the selected suitable denoising method can effectively suppress the environmental noise in the signal and improve the reconstruction efficiency of compressed sensing.