Research on Fire Prediction Method of High-Voltage Power Cable Tunnel Based on Abnormal Characteristic Quantity Monitoring

Abstract

The proportion of cable lines in the urban distribution network is increasing. The fire hazard of important cable channels is prominent, which has a serious effect on the safety and stable operation of the power system. In recent years, intelligent mobile inspection and fire extinguishing devices have been applied in tunnels. The determination of firepower and location is conducive to the rapid and effective fire suppression of intelligent devices. Therefore, this study proposes a fire early warning method of a high-voltage power cable tunnel based on abnormal characteristic quantity monitoring. Based on the modified model of complex pyrolysis and combustion chemical reaction, the fire development of different fire source powers and fire locations is simulated. The temperature distribution and characteristic gas concentration under different simulation conditions are analyzed. The results show that the monitoring data of temperature, flue gas concentration, and CO and CO2 concentration need comprehensive analysis to effectively reflect different fire conditions. The characteristic data set is selected and processed to form a total sample. The fire prediction model is trained and tested. The accuracy of the proposed prediction model is 92%.