Research On Fault Diagnosis Technology Of Reciprocating Machinery

Reciprocating machinery is widely used in chemical industry, mining industry, shipbuilding, automobile and railway industry. In this paper, two kinds of reciprocating machinery fault diagnosis technology including reciprocating compressor and diesel engine are studied. Mainly from the following four aspects:1. The mechanism of fracture failure of piston rod in reciprocating compressor is studied.The piston rod fracture failure mechanism was analyzed based on the energy balance criterion of crack propagation which is put forward on by Griffith. It is concluded that the vibration amplitude of the piston rod will increase when the piston rod fracture failure occurs. And the peak value of the displacement of the piston rod is used as the sensitive characteristic of the piston rod fracture. Finally, through the analysis of three piston rod fracture cases, it is verified that the piston rod fracture failure causes the piston rod vibration amplitude to increase. Based on the analysis of three piston rod failure cases, it is verified that the piston rod vibration fault can lead the piston rod vibration amplitude to increase. The peak value can be taken as the characteristic parameter for the alarm or diagnosis of the piston rod fracture fault.2. The feature extraction method of piston rod looseness fault of reciprocating compressor is studied.Firstly, the mechanical model is established by simplifying the piston rod of reciprocating compressor. Then, the self-excited frequency of piston rod is theoretically deduced based on the theory of bending vibration of elastic beam.And the existence of self-excited frequency of piston rod is proved experimentally by verifying the spectrum which is obtained by using the wavelet method to de-noise the experimental data. Finally, an example is given to prove that the self-excited frequency of piston rod can be used to diagnose the looseness fault of piston rod.3. The application of wavelet de-noising algorithm in diesel engine vibration signal is studied.First, the sine signal are simulated, and joined them in white noise to form noise signal. Then we explore the best de-noising methods of the three noise signals. The optimal noise reduction method is used to process the real vibration signal and we get the standard signal. Then, after the white noise is added into the standard signal, the optimal noise reduction method is studied.4. The application of inductive monitoring system algorithm in diesel engine fault diagnosis is studied.On the basis of the model of inductive clustering algorithm, the method of optimizing the disturbance point and the fault pattern recognition method of diesel engine based on inductive monitoring algorithm are proposed. Then using the real data of diesel engine vibration to train and test inductive monitoring system algorithm. After data validation, the model is correct for the fault diagnosis. The present research can provide a new way for the diesel engine fault diagnosis.