Research On Electromagnetic Characteristics Analysis And Diagnosis Method Of Short-circuit Fault Of Doubly-fed Wind Turbine

In recent years,facing the dangers of depletion of traditional energy sources and environmental pollution,wind energy,as one of the fastest-growing renewable new energy sources,has received extensive attention.In the wind power generation system,the doubly-fed wind generator has many advantages such as high utilization rate,small size,light weight,and two-way energy flow,so it has been widely used.Due to the working environment and its own structure,the failure rate of doubly-fed wind turbines is very high.Among them,stator winding faults account for a large proportion,so it is necessary to study the characteristics and diagnosis methods of short-circuit faults in the stator windings of doubly-fed wind generators.The 5.5kW doubly-fed wind generator is used as the research object.The field-circuit coupling method is used to establish a mathematical model of the doubly-fed wind generator stator winding inter-turn short-circuit fault,fully considering the cogging effect,ferromagnetic material saturation and eddy current effects and other factors.In order to truly and effectively reflect the influence of the stator winding inter-turn short-circuit fault on the electromagnetic characteristics of the motor under actual grid-connected conditions,the characteristics of the air gap flux density change under the single-phase inter-turn short-circuit fault of different degrees are analyzed,and the variation characteristics of the amplitude and phase difference of the three-phase current on the stator side,as well as the variation characteristics of the amplitude and frequency of the electromagnetic torque component,to provide more references for the fault diagnosis of the doubly-fed wind turbine.FFT and wavelet packet transform are used to analyze and process the stator current signal,and the harmonic law of the current value before and after the fault and the energy transformation law of different frequency bands are obtained;in order to identify and judge the fault category,an artificial intelligence quantitative method is proposed —BP neural network,based on FFT and wavelet packet transform analysis of the current feature quantity as a sample input into the neural network,the output mode is the expected fault type,and then the network layer is learned and trained separately,through the number of training times,the two algorithms are compared in terms of error size and fit,and then the BP neural network is better applied to the diagnosis of short-circuit faults in the stator windings of doubly-fed wind turbines.