基于模态分解的载流管路振动线谱激励载荷类型辨识研究OA

To accurately identify the excitation load types of line-spectrum,starting from the vibration theoretical model,the probability distribution characteristics of the correlation functions of pipeline vibration responses under discrete-spectrum and continuous-spectrum excitations were derived,and a new method for identifying whether pipeline vibration line spectra were excited by discrete-spectrum or continuous-spectrum loads was proposed.This method utilized the variational mode decomposition(VMD)technology to decompose the autocorrelation function of multi-frequency line-spectrum vibration responses into single-frequency intrinsic mode functions,and their probability distribution characteristics were obtained.By leveraging the differences in probability distributions of vibration response autocorrelation functions under different load excitations,correlation analysis methods were employed to accurately identify the excitation load type of each vibration line spectrum.This method relied solely on measured vibration response data,and was characterized by high efficiency and reliability.The effectiveness of the method proposed was demonstrated through an experimental model of a U-shaped pipe conveying fluid under pump excitation.Results show that this method can accurately identify no less than five types of vibration line spectrum excitation loads.