The Research Of The Early Vibration Fault Identification Technology Of Steam Turbine Based On Wavelet Analysis

In recent years, electric power industry gets fast development, power generation capacity amount rise year by year, electric power demand tension still dominate the most region, and some of newly-built units are being in commissioning and startup crucial stage, units safety and economic operation is the basic guarantee of power generation, therefore the vibration monitoring of steam turbine and fault diagnosis technology have been put forward higher requirement. This paper revolves around the early vibration fault identification technology of steam turbine to spread out research, applies the wavelet analysis theory on the rotor vibration signals of steam turbine to carry out thorough analysis, especially aims at weak signal feature that of early vibration signal, digs the identification detection technology suited early vibration signals, establishes foundation for early vibration fault diagnosis.This paper is taking wavelet theory and its application as foundation, adopts multi-resolution analysis and wavelet package decompose thought. With the common rotor vibration fault of steam turbine as research object, bring forward the methods based on discrete wavelet transform to thin frequency interval, draw the shaft center orbit picture after wavelet de-noise, carry out energy spectrum analysis after wavelet package decompose. Then applies it on the simulated rotor test-bed , testified with real data, have gotten good effect. Have offered beneficial way of thinking for research of the large steam turbine units condition monitoring and early fault diagnosis technique.