Investigate and Experimental Verification of Early Monitoring and Warning Fire Extinguishing System for Substation Switchgear Cabinet Fire

Abstract One new fire warning technology for substation switchgear cabinet is proposed in this paper, which implements real-time monitoring of key characteristic parameters such as temperature, smoke concentration, combustible gas concentration, and flame state before a fire occurs. The experimental data were systematically trained and validated by using neural network algorithms. The switch cabinet integrates fire characteristic parameter detection modules, which use ZigBee wireless communication network to transmit the collected data to the central concentrator. Subsequently, a pre trained neural network model is used to analyze and recognize the received fire feature data. Once signs of fire are detected, the fire extinguishing mechanism is triggered to release perfluorohexane for rapid extinguishing. To verify the effectiveness of the system, we used the KYN28A-12 switchgear cabinet as a platform to construct a real early fire monitoring, warning, and automatic fire extinguishing system, and conducted rigorous experimental evaluations of its performance. The experimental results show that the system exhibits high accuracy and reliability in fire detection, and can achieve extremely early warning of switchgear cabinet fires, providing a solid technical foundation and guarantee for the fire safety protection of substations.