On-line Detection Of Fault Arc And Research On Intelligent Circuit Breaker In Low Voltage Distribution System

In the family distribution line,the phenomenon of fault arc is easily caused by the aging of insulation,or the loose connection of electrical equipment.The generation of arc fault is accompanied by a lot of heat,which can easily cause inflammable and explosive materials to burn or even explode,causing life and property losses to the people.Studies have shown that even only 2 to 10 A of the arc current can cause high temperatures up to thousands of degrees.Therefore,the principles and characteristics of the fault arc have been extensively studied,so that we can timely and accurately determine whether there is a fault arc in the circuit,thereby cutting off the circuit and ensuring the safety of the people,which has strong theoretical and social value.This paper refer to the International Electrotechnical Commission IEC6206-2013 Ed1.0 MOD "electric arc fault detection(AFDD)of the general requirements",and according to the national standard GB14287.4-2014 "arc fault detection device" and GB/T 31143-2014 "arc fault protection appliances(AFDD)of the general requirements of" design series of point contact arc fault test system.Based on this set of experimental platform,signal acquisition of the electrical load,were collected on the normal operation and current of arc fault occurs,the fault features,focusing on analysis of different load current waveform characteristics in two conditions for the extraction and quantification and classification method.The main contents of the paper include:(1)The origin and process of arc are studied from two aspects of physics and circuit theory.The development status and background of fault arc detection device are introduced.Based on the differential equation of arc gap energy balance,a series fault arc simulation model is built on MATLAB/Simulink,and the voltage and current waveform of the simulation arc is extracted by using the relevant modules,and the waveform is compared with the measured waveform.(2)Experiments were carried out to study the different load to obtain different normal load and current data series arc fault condition by the fault arc test platform,the data analysis and comparison,find out the characteristics of series arc fault,summarizes thetypical characteristics of arc fault current signal in time domain.(3)Introduce a wavelet packet denoising and EMD based fault detection algorithm based on fault arc,arc fault accurate identification circuit,and data were extracted using the experimental platform,based on the signal denoising of wavelet packet decomposition principle,through EMD to get the feature vector for SVM training,so as to realize the arc fault detection.(4)From the two aspects of the requirement and technology of the arc detection device,the overall design framework of the hardware and software is designed,and the selection of hardware and the software environment are introduced.