Research On Mathematical Model And Diagnosis Methods Of Series Arc Fault In Mine Power Supply System

Due to poor contact and other reasons,series arc faults may occur in the electrical connection point of the mine power supply system.Series arc fault with certain latent and hidden period is one of the main causes of electric fire in mine power supply system.Because it is difficult to carry out series arc faults experiments in mine power supply system,obtaining the typical characteristics and diagnostic methods of series arc faults has become the key of the mine fault circuit breaker.Firstly,series arc faults experiments are performed in AC 36V,AC 110V,AC 220V and AC 380V.Then,the mathematical model parameters of series arc faults are obtained under different experimental conditions by the numerical analysis.Based on the correlation between circuit parameters and mathematical model parameters,the mathematical model of series arc faults is established.The validity of the model is verified by the simulation analysis of series arc faults of the experimental system.Next,the feature parameter database of series arc faults in mine power supply system is established by simulating and analyzing series arc faults and other circuit faults in mine power supply system.Finally,the diagnosis model of series arc faults in mine power supply system based on the K nearest neighbor algorithm is proposed by comprehensively comparing the accuracy,fastness,anti-jamming performances of the diagnosis model.The research results are of great importance to prevent and control the electrical safety accident caused by series arc faults and ensure the reliability of power supply in mine power supply system.The main conclusions of the thesis are as follows:（1）There is a large correlation between the circuit parameters of voltage rms,current rms,power factor and the mathematical model parameters?_m,?,_sp,?,g₀ of series arc faults.The neural network black box model can be used to predict the parameters and establish the mathematical model of series arc faults in mine power supply system.（2）The established mathematical model of series arc faults can be used to implement the feature analysis and fault diagnosis of series arc faults in mine power supply system.（3）The number of zero-crossing points,peak-to-peak value,variance,kurtosis coefficient and margin factor of series arc faults currents can be used as the time domain characteristic parameters of series arc faults.The harmonic distortion rate and frequency variance of unilateral power spectrum after trap-wave processing of series arc faults currents can be used as the frequency domain characteristic parameters of series arc faults.The energy entropy and peak-to-peak value of wavelet packet coefficient of series arc faults currents after trap-wave processing can be used as the time-frequency domain characteristic parameters of series arc faults current signals.（4）The diagnosis model of series arc faults based on the K nearest neighbor algorithm has an obvious advantage in the accuracy,anti-load current fluctuation disturbance,anti-environmental noise disturbance,which can be used to diagnose series arc faults in mine power supply system.（5）The diagnosis model of series arc faults based on the K nearest neighbor algorithm in coal mine power supply system has the function of diagnosis and phase selection of series arc faults.