Journal of Shanghai Jiaotong University ›› 2018, Vol. 52 ›› Issue (12): 1618-1626.doi: 10.16183/j.cnki.jsjtu.2018.12.011

Previous Articles     Next Articles

Separation of Single Channel Radio Mixed Signal Based on Variational Mode Decomposition

JIANG Chundong,WANG Jingyu,DU Taihang,HAO Jing,LONG Chao   

  1. 1. School of Control Science and Engineering, Hebei University of Technology, Tianjin 300130, China; 2. Control Engineering Research Center of Hebei, Tianjin 300130, China; 3. Department of Computer Application, Shijiazhuang Information Engineering Vocational College, Shijiazhuang 050000, China

Abstract: Aiming at the separating difficulty of single channel radio mixed signal and low accuracy of its separation under complex electromagnetic environment, a separating method of single channel radio mixed signal by using twice variational mode decomposition (VMD) algorithm is proposed. Firstly, the VMD algorithm is used to roughly separate single channel radio mixed signal for the first time. And the VMD algorithm is compared with the ensemble empirical mode decomposition (EEMD) algorithm. The results show that the separated signals by the former are better in the time domain, frequency domain, signal to noise ratio and similarity coefficient than those obtained by the latter. Then, the parameters of variational mode decomposition algorithm are optimized by using the improved quantum particle swarm optimization algorithm, thus the number of components and the value of penalty factor are determined. And then, the optimized parameters of VMD algorithm are used to fine separated signals from VMD algorithm. The numerical simulation results and the experimental signal analysis results show that the signal accuracy of the second separation is higher than the first one. And it is proved that the proposed algorithm is valid for separating the single channel radio mixed signal.

Key words: variational mode decomposition (VMD), ensemble empirical mode decomposition (EEMD), improved quantum particle swarm optimization algorithm

CLC Number: