Exploration of Electric Vehicle Charging State Monitoring and Fire Early Warning Based on Neural Network

As an important manifestation of new energy utilization, electric vehicles have widely penetrated people's daily lives. However, the frequent occurrence of fire accidents during its charging process has also become a key issue restricting the development of electric vehicles. Based on CNN and BiGRU, this paper constructs a charging state monitoring and fire warning model by making full use of electric vehicle charging historical data and verifies the prediction accuracy and the pre-alarm threshold of the model through simulation experiments. The experimental results confirm that the technical approach adopted in this paper can realize the real-time monitoring of the charging process of electric vehicles, TIMELY detect faults, and issue pre-alarm signals, which provides a good guarantee for the charging safety of electric vehicles and fire early warning.