基于GAHE-VMD与SVD-SCEC的管道信号联合去噪法

Pipeline Signal Complementary Denoising Method Based onGAHE-VMD and SVD-SCEC

为了更有效分析管道信号,提出一种基于采用在恶劣环境下的遗传算法(Genetic Algorithms in Harsh Environments,GAHE)优化变分模态分解(Variational Mode Decomposition,VMD)联合奇异值分解(Singular Value Decomposition,SVD)与选择性累计能量贡献率(Selective Cumulative Energy Contribution,SCEC)的互补去噪方法。首先,提出用GAHE算法优化VMD算法并结合相对熵对信号中的中高频噪声进行初步消噪,解决VMD参数难以确定和传统遗传算法收敛慢的问题。其次,提出采用SCEC算法结合SVD算法对信号中残留的中低频噪声进行消噪,解决非线性、非平稳信号中大数量级的直流分量影响奇异值选择的问题。最后,通过实验与分析表明:GAHE优化算法收敛速度更快;SCEC奇异值选择法的抗直流能力更强;所提算法的处理效果较优且算法两部分具有互补特性。

In order to analyze pipeline signals more effectively,a complementary denoising method based on combination of GAHE(Genetic Algorithms in Harsh Environments),optimized VMD(Variational Mode Decomposition),SVD(Singular Value Decomposition)and SCEC(Singular Value Decomposition)is proposed.Firstly,the GAHE algorithm is used to optimize the VMD algorithm and the relative entropy is used to eliminate the middle and high frequency noise of the signal,which can solve the problems of the difficulty of VMD parameter determination and slow convergence of traditional genetic algorithm.Then,The SCEC algorithm combined with the SVD algorithm is used to denoise the residual low and medium frequency noises in the signal,which can solve the problem that the large-magnitude DC components of nonlinear non-stationary signals interfere the selection of singular values.Finally,through experiments and analysis,it is concluded that the proposed GAHE optimization algorithm has the advantage of faster convergence,the SCEC method has stronger anti-DC capability,the proposed algorithms have better processing effect,and both parts of the algorithm have complementary fearures.