Low Voltage Series Fault Arc Detection Based On Power Spectrum Estimation

In low-voltage distribution network,series fault arc is a kind of the most harmful and most common electrical fault.Due to the fault arc is not easy to be timely and accurate detection and discovery,it is a huge potential threat to the power network.Therefore,the study of the low voltage series fault arc detection method has very important practical significance.In this paper,a series of fault arc is used as the research object,and a thorough research on it from theoretical calculation,arc collection analysis and software model simulation.The theoretical of the fault arc is analyzed in detail.Based on the theory of gas discharge,the generation process of arc is analyzed in detail from the microscopic point of view.To determine the generation and classification of the fault arc.By solving the inductive series arc fault circuit,the relationship between arc voltage and circuit power factor is obtained,and the basic characteristics of arc in series are summarized.A experimental platform of low voltage series arc fault arc is established and the experimental study is carried out.The establishment of low voltage series fault arc experiment platform based on the newly issued national standard.Experiments are carried out by using different categories of linear loads and nonlinear loads in series.The experimental data is collected and sorted out.The waveform of experimental current and voltage are analyzed,and the characteristics of the series arc are verified.Based on power spectrum estimation to research on the detection method of series faults arc.The power spectrum estimation of the experimental current data is made by using the method of the average amended periodic table,and the power spectrum of the non arc and the arc is obtained.Extract power spectrum characteristic quantity,establish experimental data sample storehouse.The support vector machine is used to classify the data samples.The correct rate of fault arc detection is 95%.And this result has achieved the desired results.Software simulate is completed in detection process of series arc fault in secondary distribution system.The transmission of the signal in the low voltage power line is analyzed,and the transmission matrix model is derived.In the SIMULINK simulation system,the fault arc detection is simulated and analyzed in secondary distribution system and the simulation current data is collected.The power spectrum estimation of the simulated current data is made by using the method of the average amended periodic table,the simulated power spectrum of the non arc and the arc is obtained,and the power spectrum characteristic quantity is extracted.