Abstract: Electrical fire monitoring and early warning are of great significance for fire prevention and control. In response to the complexity of early signs and the limited accuracy of risk prediction, this paper analyzes the research progress on the coupled mechanisms of multiple electrical faults which may lead to electrical fires, including overheating, short circuits, arc faults, overloads, and glowing connections, further compares fault identification methods and models, and provides a systematic review of the state-of-the-art in electrical fire monitoring and early warning technologies. The review highlights the critical scientific issues and technical challenges from the viewpoints of failure mechanisms, monitoring and identification technologies, and early warning systems. It is proposed to carry out experiments on reproducing disaster scenarios involving the transformation between different electrical faults, thereby identifying characteristic parameters of early warning signs associated with primary fault sources of electrical fires. Furthermore, a predictive model for overheating risk in electrical circuits was developed by integrating multi-source information, including electrical parameters, pyrolysis gases, and environmental factors, enabling accurate early-stage warning of overheating conditions. Multi-source recognition and localization approach for glowing connections that integrates electrical, thermal, gaseous, and optical signals was developed, and a monitoring strategy for fault arcs under multi-factor coupling conditions was proposed. Research on an intelligent monitoring and early warning system for electrical fire hazards based on cloud-edge-device collaboration, the construction of an electrical fault disaster spectrum, and AI-powered electrical fire early warning agents are key trends in improving the accuracy of monitoring and early warning.

Key words: electrical fire, early signs, monitoring and early warning, intelligentization