Resistive Leakage Current (Igr) Determinant Analysis Using Dynamic Regression Analysis

Abstract

In electrical fires, resistive leakage current (Igr) is considered as the main cause of electrical safety accidents such as electric shocks and fires. In this study, the determinants of the Igr were empirically analyzed using the electrical data collected in real time using the IoT sensor. An 1 lag Igr value (Igr data 1 h ago) was observed to significantly explain the current Igr. In particular, the influence of the 1 lag Igr value was weakened in the risk stage rather than the non-risk stage, whereas the influence of other factor variables was strengthened. This can be regarded as a dangerous sign of an electric fire, and the results of this analysis can be used to identify and predict the prognosis of future fires. These results show that the spread and prognosis of a fire can be understood by dividing it into risk and non-risk stages.