Application Of Compressed Sensing In Remote Mechanical Fault Diagnosis

Mechanical failure exists in any mechanical system,any kind of failure will be the normal operation of machinery and equipment play a role in hindrance.especially the impact on the safety of enterprises.The damage caused by mechanical failure is immeasurable.Therefore,the inspection of mechanical failure is essential.The existence of any detection method has its scientific significance.Efficient and objective methods of fault detection are constantly updated with the development of technology.In 2006,Candes introduced a new signal sampling theory at the International Conference of Mathematicians.It is pointed out that as long as the number of samples required by the traditional Nyquist sampling theorem is far less than that of the traditional Nyquist sampling theorem,the original signal can be reconstructed accurately or with high probability.In this paper,a novel method of compressed sensing is used to detect and analyze the inner ring failure,outer ring failure,and rolling element failure,as well as axial piston pump slide fault,loose shoe failure and shoe removal fault.The method first compresses the fault data.greatly reducing the storage space for data,improve the transmission efficiency of the data.Thereby realizing the remote transmission of the compressed data.The transmitted data are recovered by the orthogonal matching pursuit algorithm.Thus the fault signal can be extracted efficiently.Through the simulation of the data analysis can show that compressed sensing method is a scientific and effective method of fault diagnosis.In this paper,the compressed data achieve a remote transmission in the VC ++ language environment,the transmission of compressed data recover through the orthogonal matching tracking algorithm,through MATLAB simulation of the original fault data and compression sensing method to reconstruct the data of the time domain waveform The contrast signal and the envelope signal of the original signal and the reconstructed signal are analyzed,and the time-domain waveform of the reconstructed signal and the original signal is analyzed.The result of the reconstructed signal is controlled by selecting different compression quantities.In the error range,the envelope demodulation power spectrum curve of the reconstructed fault signal and the original fault signal can accurately diagnose the various faults.The experimental results are authenticity and reliability,which verifies the method in the remote mechanical fault diagnosis Domain application feasibility.