Research on fire early warning index system of coal mine goaf based on multi-parameter fusion

Abstract

AbstractIn order to improve the precision of goaf fire early warning outcomes, this paper obtains the temperature characteristic index of goaf fire early warning by using a coal spontaneous combustion thermogravimetric test and a coal spontaneous combustion programmed heating test. The major gas index and auxiliary gas index for early warning are derived using the integration of the Graham coefficient and grey correlation approach. The D-S evidence theory, which involves optimizing weight allocation, is utilized to integrate the early warning temperature index and various gas indexes. Based on the fusion results, a coal mine goaf fire early warning index system is developed through multi-parameter fusion. The early warning index system is then validated through a programmed heating experiment. The results show that the process of coal spontaneous combustion can be categorized into six distinct stages: latent stage, oxidation stage, critical stage, pyrolysis stage, fission stage, and combustion stage. These stages are determined by the characteristic temperatures of coal spontaneous combustion, which are 31.7 °C, 54.8 °C, 153.7 °C, 204.5 °C, and 241.6 °C. The major gas index for early warning of goaf fires can be determined by 100∆(CO)/∆O2(%). Additionally, auxiliary gas indexes such as C3H8/CH4, C3H8/C2H6, C2H4/C2H6, and C2H2 can be used for goaf fire early warning. The programmed heating experiment shows that the early warning system software is designed by the multi-parameter fusion goaf fire early warning index system is accurate and effective. The selection of the goaf fire early warning index is more rational and precise when using the multi-parameter fusion goaf fire early warning index system based on the D-S evidence theory of weight allocation. It offers robust support for enhancing the goaf fire early warning index system and predicting coal mine goaf fires.