Series Arc Fault Detection And Feature Extraction

The arc fault is discharge of electricity phenomenon across an insulating medium(such as: air),between two electrodes.The partial electrodes will volatilize and heat distribution usually when arc occurs.Arc fault can distribute lots of heat,which can cause fire,if the power can't cut off in time.The fact provided that the fire caused by arc fault lead to many people lose their lives and property.How to effectively avoid fire caused by arc,and keep the people's lives and property safe,is now urgent need to solve the problem.In order to timely and effectively detect whether an arc fault appears,the arc fault feature signal need to be extracted,and then get on arc fault eigenvalues.Use eigenvalues to detect arc fault.In order to more easily collect fault arc data,arc fault simulation experiments will be adopted for arcing signal acquisition.Experiments in accordance with the relevant requirements of the reference standard,Four kinds of circuit will be designed with Seven kinds of shielded load.Finally dozens group of arc data were be obtained.Arc fault feature quantity in time domain were obtained by comparing dozens group of arc data.For more fully grasping the characteristics of arc fault,fast Fourier transform analysis was used to get arc fault feature in frequency domain.Arc fault sudden change feature also can be get with wavelet transform.At last,Three characteristics of arc fault(such as: The rate of increase is relatively quickly,Different positive half cycle duration,Differences sample values of positive half cycle)will be adopted in software,Use STM32 MCU to make sure arc fault detection.