With the rapid development of the electric vehicle industry, charging piles, as critical infrastructure, face increasingly prominent electrical fire safety hazards. To address the issues of low accuracy and slow response in traditional warning solutions, this paper designs an electrical fire warning system for charging piles based on Internet of Things (IoT) technology and hybrid algorithms. The system adopts a three-layer architecture of perception layer, network layer, and application layer. It collects electrical and environmental parameters through high-precision sensors and utilizes LoRa, NB-IoT, and 5G technologies for reliable data transmission. Innovatively, a hybrid intelligent warning model integrating 1D-CNN, LSTM, and Random Forest is proposed, effectively capturing transient abnormalities and long-term trends in electrical signals. Experimental results show that the system’s current monitoring error is less than ±0.1%, the warning accuracy reaches 95%, the response is rapid, and it exhibits excellent robustness and environmental adaptability in high/low temperature and strong electromagnetic interference environments, providing reliable technical support for the safe operation of charging piles.