Application of Wavelet Packet Analysis in Turbine Fault Diagnosis

Abstract

Experimental platform is used to simulate typical faults of turbine. Based on the frequency domain feature, energy eigenvector of frequency domain is presented in the wavelet packet analysis method, and the way of best tree is used to choose symptom. Finally, the fault states are recognized using neural network, and the simulations show that it makes a good performance with the method.

Research of Fault Feature Extraction Based on High Order Cyclic Statistics for Reciprocating Compressor Gas Valves

“…As the complex structure and more excitation sources, the vibration signals of reciprocating compressor often represent non-stationary and non-Gaussian, thus it is difficult to diagnose the faults of reciprocating compressors [1]. At present, approaches of the fault diagnosis for reciprocating compressors are all based on the assumption that the vibration signals are stationary [2].…”