Based On Wave-let Transform Single-phase-line-open Fault Selection And Location Of Distribution System

With the improvement of automatic level in distribution systems,the single-phase-line- open fault selection and location is becoming the focus which is paid attention to and investigated.the study about diagnosis of the fault is urgent daily,too.After analysis,there is great change of positive-current component when the small current grounding power systems happen single-phase-line-open fault.The transient component of positive-sequence current is marked as the occurrence of the fault.And it can be concluded the magnitude and phase characteristic of it,which is also bigger than or equal to the negative-sequence current component.Through much simulation and analysis,there are plenty of harmonics in the fault current in the transient course.And it is mutational.Based on this,the paper uses wave-let transform to obtain useful fault transient components,which is used as a novel tool of signal analysis.A new criterion of fault line selection is constructed based on wave-let transform modulus maxima principle of singularity detection.Which is combined with BP(Back Propagation) neural network for fault location to realize the mapping relationship between modulus maximum and fault point location.The criterion takes advent- age of wave-let disposing the transient singular or faint signal,and makes up for the lack of the traditional analyzing principle based on steady signal and enhances the anti-interference ability.So it has higher reliability and accuracy.This article has deduced theoretically characteristics and changes of zero-sequence- voltage,negative-sequence and positive-sequence current in fault line of the three types of single-phase-open-line faults in distribution network by the arc-suppression coil.And it has simulated the condition of fault selection in it using the plan.This will be of great significance to improve the distribution network automation.According to the characteristics of wavelet transform modulus maxima and BP neural network,it has constructed wavelet neural network model,has firstly brought forwand a new plan for fault loation to combine the modulus maxima of positive-sequence current with BP neural network.The ATP and MATLAB softwares are used to simulate the 10kV lines distribution network.And the results show the reliability and accuration of the method of the selection and location.