An Experimental Study of Sensor-less Fault Diagnosis on Rolling Bearing of Wind Turbine Generator System

Abstract

This paper study sensor-less fault diagnosis on rolling bearing of wind turbine generator system, by means of establishing an experiment that can simulate doubly fed wind power generation system, on basis of the mathematical model that reflects magnetic flux density impacts on the electrical signal, using the method of Motor Current Signature Analysis. According to collect the electrical signal and shaft's made by the generator bearing that works in normal and fault condition, analyzing the fault principle for the bearing of the transmission chain and the fault signal's feature. The experiment results show that the stator current signal's feature is familiar with the predicted failure frequency and the characteristic frequency of the main shaft frequency signal When rolling bearing fails, and provide the detecting sensor-less method of rolling bearing wind generator fault detection and diagnosis is feasible as well.