| In practical engineering applications,there are many complex impact phenomena,most of which are detrimental to mechanical operation.By analyzing the impact signal,the characteristics of the impact can be obtained,mechanical faults can be judged,the cause of the fault can be obtained,and corresponding measures can be taken to reduce the harm caused by the impact.At present,impact signal extraction technology is widely used in fields such as mechanical fault detection,by extracting feature information from signals,in order to more effectively identify and predict mechanical faults.It can help analyze a large amount of data,more accurately detect mechanical faults,and improve the reliability and safety of mechanical equipment.At present,acceleration sensors are mainly used to extract impact signals.However,due to the complex on-site environment,the raw data collected by sensors often has drawbacks such as noise interference,low sampling frequency,and poor signal-to-noise ratio,which greatly affects our diagnosis of the actual situation and is difficult to meet the needs of engineering analysis.Moreover,the obtained impact signal is a time-domain signal with only a small number of physical parameters such as peak values,which makes our understanding of the signal insufficient to make accurate judgments.The classical decomposition method can decompose shock signals to obtain multiple characteristic shock waveforms,each with accurate physical parameters,including maximum acceleration,angular frequency,damping,initial time,peak time,and phase.But this method processes the impact signal in the time domain,which will generate cumulative errors and affect the accuracy of signal extraction.In response to this issue,this paper proposes a method of using reconstructed shock signals to replace the original shock signals in complex environments,namely the shock signal decomposition and reconstruction method.This method overcomes the cumulative error caused by successive subtraction in previous signal decomposition;At the same time,it can also characterize the impact signal,decompose and reconstruct the impact waveform,and the components of the reconstructed impact signal can also be represented by physical parameters,including maximum acceleration,angular frequency,damping,initial time,peak time,and phase,with clear physical significance.The specific content is as follows:1.Theoretical research on frequency domain decomposition and reconstruction methods for impact signals.Introduced the relevant theories of single degree of freedom systems,studied the characteristics and physical indicators of impact signals,including the mathematical expression of single frequency signals,and theoretically analyzed the mathematical expression of impact signals.Research on the preprocessing theory of impact signals under strong background noise,including digital filters and wavelet decomposition and reconstruction methods.The Fourier transform of the signal is studied.According to the superposition of the signal in the linear system,the idea of replacing the original shock signal with the reconstructed signal is proposed,which lays a theoretical foundation for the followup work of this paper.Studied the theoretical basis for calculating the shock response spectrum and signal energy,in order to evaluate the effectiveness of reconstructing the signal.2.Numerical simulation analysis of the frequency domain decomposition and reconstruction method for impact signals.According to the frequency domain decomposition and reconstruction method of impact signals proposed in this article,numerical simulation analysis is conducted on the constructed simple ideal signal and the simulated impact signal of a certain special vehicle when subjected to impact,to verify the feasibility of this method.And set up comparative simulation analysis,that is,use classical decomposition method to conduct numerical simulation analysis on the same impact signal,and evaluate the difference between the reconstructed signal and the original signal by calculating the impact response spectrum and signal energy.The simulation results show that for a simple ideal signal,the reconstructed signal obtained by the frequency domain decomposition and reconstruction algorithm of the impact signal accounts for 91.3% of the original impact signal energy,while the reconstructed signal obtained by the classical decomposition method accounts for 90.5%of the original impact signal energy.The simulation results of the shock response spectrum indicate that the frequency domain decomposition and reconstruction method of the shock signal extracts better information than the classical decomposition method in the low frequency range of the signal.For the simulated impact signal of a special vehicle under impact,the energy ratio obtained by the frequency domain decomposition and reconstruction method of the impact signal reaches 90.1%,indicating that the frequency domain decomposition and reconstruction method of the impact signal proposed in this article is also applicable to the decomposition and reconstruction of complex signals and has a wider application. |
