Neural Network Technology for Electrical Fire Early Warning System

Abstract

Electrification not only promotes the rapid progress of productivity and human civilization, but also provides a greater possibility for the occurrence of fire. Electrical fire is mainly caused by abnormal high temperature caused by line overload, short circuit, poor contact, arc spark, electric leakage, lightning or static electricity, which leads to ignition of surrounding combustibles or spontaneous combustion of cables. At present, the detection of electrical fire is mainly through the detection of the surface temperature of various wires and cables and the detection based on the electromagnetic principle, which has the problems of high false alarm rate and difficult maintenance. Considering that most electrical fires are caused by abnormal high temperature of the line due to electrical faults, which makes the insulating layer self ignite or ignite the surrounding substances, while the insulating layer made of plastic or rubber has the characteristics of high-temperature thermal decomposition and will release specific gas when working above the rated temperature, this paper proposes the technical research of neural network on the electrical fire early warning system, The application research of neural network technology in electrical fire early warning system is the application research of artificial intelligence and machine learning technology in fire detection. Its main purpose is to use artificial intelligence to detect and predict the occurrence of building fire, and then provide an early warning means to prevent people from fire accidents.