摘要：For selective detection of earth faults in medium voltage networks, the general protection can be used, which unlike individual protections controls earth currents in all lines supplied from the substation. To detect high-resistance and arc faults, it is possible to use in such protections the details from wavelet decomposition of earth currents and the Bayesian conditional probability criterion. The paper presents a protection model using details of earth currents and the Bayesian criterion. Significant parameters of the model were determined: sampling frequency, wavelet types and decomposition levels, parameters of the probability density function and decision conditions. The effectiveness of the protection in the case of non-stationary high-resistance short-circuits in the network with natural asymmetry and harmonic distortion was also assessed.
其他摘要：For selective detection of earth faults in medium voltage networks, the general protection can be used, which unlike individual protections controls earth currents in all lines supplied from the substation. To detect high-resistance and arc faults, it is possible to use in such protections the details from wavelet decomposition of earth currents and the Bayesian conditional probability criterion. The paper presents a protection model using details of earth currents and the Bayesian criterion. Significant parameters of the model were determined: sampling frequency, wavelet types and decomposition levels, parameters of the probability density function and decision conditions. The effectiveness of the protection in the case of non-stationary high-resistance short-circuits in the network with natural asymmetry and harmonic distortion was also assessed.
