Research on fault arc detection method based on multimodal fusion strategy

With the increasing number of electrical fires caused by fault arc, effectively identifying the fault has become a very meaningful thing. According to actual product testing, it has been found that many fault arc detection products on the current market have high level of both false positive rate (FPR) and false negative rate (FNR). A high FPR refers to frequent misleading alarms, which may lead to unexpected power outages. And a high FNR means that fault arc has occurred but the system has not detected it, which may lead to a fire. The contribution of this study is the proposal of a multimodal fault arc recognition algorithm, which effectively reduces the FPR and FNR of the arc detection system, and helps to improve the comprehensive performance of the electrical safety warning system.