The Research Of The Fault Diagnosis For The Pump Vales Of Reciprocating Pump

Reciprocating pump is a widely used machine equipment in engineering. It is rather difficult to monitor the state of reciprocating pump owing to its complex construction and so many stimulation sources. Although the researches have been made and some achievements have been gained, but practical application is insufficient, and less in the application of industrial devices. At present, two key issues of fault diagnosis for the pump valves of reciprocating pump are extracting the fault feature information of stationary process efficiently from system feature signals and identifying the specific faults correctly with analysis of causes. The research work which is combined with practice in this thesis is based on predecessors' achievements, and based on vibration signals of pump vales of reciprocating pump , the fault feature information is extracted, intelligent diagnosis methods are adopted and fault diagnosis is effectively solved for reciprocating pump valves.This paper uses ordinary vibration signals as system feature signal to pickup fault characteristic vectors. The advantage of this method is its vibration signals are easy to gain and analyze. Then the method of time domain analysis and wavelet packet analysis is used to extract the fault feature information which is obviously different for different fault. To advance the accuracy of the fault diagnosis, time domain analysis and wavelet packet analysis are intermixed to gain fault characteristic vectors. The paper constructs a three-layer forward neural network to diagnosis the fault and trains the network with characteristic vectors extracted, at the same time adapt the initial data of the network through the test. After confirm the construction and parameters of the net, the result is valid and justified through the verification using the test sample data. When the neural network is used to diagnose the fault of the pump vales, it is found that the methods can raise accuracy.